ASRNews

Market, Investment and Technical News of the Emerging Speech Technology Industry

INSIDE THIS ISSUE

The Emperor is Naked!

Sterling Audits Usability Work: The work that is being done by Sterling Audits is unique. It is gathering data on the usability of existing voice response systems using standardized methods. The methods that are used for evaluating the usability of a voice response system are provided which makes it relatively easy to compare the usability of a particular system to others. Previously, I’d run into a poorly implemented system and simply conclude that it sucked. Now, with the Sterling Audits technology, I can determine how badly it sucks relative to other voice response systems. You can even determine just what is causing the voice response system to suck, and fix it. Kudos to Sterling Audits!

Bizarre iVoice Future Plans. Essentially getting out of the product business and focusing on the IP extortion business. Selling off the speech technology assets. All sorts of weird transactions that are impossible to make any sense of. A financial shell game! Appears to be intended to impress the pump-&-dumpers that are generally ignorant. The ASR that iVoice has, was obtained from Entropic prior to them being acquired by Microsoft. I would recommend that any potential purchaser request a copy of the ownership title. What they have is ancient technology that has virtually no support. From what I’ve seen, it doesn’t work very well. A bit of research on the iVoice customer list would quickly lead one to the conclusion that many of the customers that iVoice claims, do not really exist as iVoice customers. Patent portfolio appears to consist of patents that were either fraudulently obtained and/or have little commercial value. Finding out where the milk is located at the supermarket is not a big problem for most people and a system that lets you locate the milk via telephone is solving a non-problem. Obtaining patents in the talking container area without having done any work in this area, is shaky. Appears to primarily running with other peoples ideas. iVoice stock has been tubing during the last year. Even with 5 billion shares outstanding, market cap is just a few million $s. CEO keeps selling iVoice stock. Cute business while it lasts – keep hyping, issue more stock and sell it. Like printing money, and similarly illegal. When the SEC notices this, it is gone.

It’s the caller liking the voice, stupid! The speech industry methods that are used to determine the appropriate sounding voice are ineffective. They are extremely complex and result in not seeing the forest for the trees. I recently saw an evaluation/assessment of the voice characteristic of an automated system. The voice characteristics that were looked for were; Chatty-sounding; professional-sounding; friendly-sounding; cold-sounding; bored-sounding; bossy-sounding casual-sounding; arrogant; commanding; patronizing; shy; fawning; pleading; mean; or silly. Making it more complex, each of these were assessed as to whether they sounded like either a machine or a person. Despite the large number of factors that were considered, the only important one was ignored. The only aspect of the voice that is important is whether the caller likes it or not. Other characteristics of the voice are really secondary as long as the callers like it. A caller not liking the voice is distracted. Last November, I had a note in my newsletter about Q-Scores: Entertainment Industry Knows How to Select Voices: For decades, the broadcast TV industry has been using a very simple mechanism to select the news anchor. It is called the Q score. Virtually every local news program looks alike. Based on extensive research, it has been found that the basic format works equally well from Bangor, Maine to San Diego. Since the way in which you present news, weather and sports is essentially a commodity, what matters most is what the vehicle that delivers it is - the anchor person. Determining the Q Score is painfully uncomplicated. It puts a potential anchormen or anchorwomen in front of a test audience and asks for just one of two responses: I like or I don't like. The Q Score technique seems like a wonderful way to select an appropriate voice for an application. Ask a hundred people to listen to a voice and indicate whether they like it or not. The one that gets the most "likes" has the highest Q Score and wins. Clean and simple approach. So much better than the expensive and subjective approach that is used by much of the speech industry, that yields such marginal results.

ScanSoft Auditor Resigns: Early this month, ScanSoft announced that their auditor (PwC) had resigned. A couple of weeks later, ScanSoft hired a new CFO. Their prior CFO left, but did not appear to have another job, which suggests that his leaving was sudden and not his choice. Neither ScanSoft or PwC are providing any sort of reason why PwC decided to resign. Naturally, they claim that no problems exist. L&H/KMPG management denied that anything was wrong. With no reason being provided for PwC bailing out, you need to assume the worst. Not likely to be a good for ScanSoft stockholders. Stock price did not take a hit. ScanSoft’s stock price has been so under water that it had no where to go.

Sterling Audits Data Shows how bad our systems are: I recently read the Sterling Audits "Voice Response Usability Almanac, 2004". The data confirmed a suspicion that I had developed from casual use of them that many of the voice response systems are really bad from the perspective of a caller. For example, the Sterling Audits’ Usability Index addressed getting to an operator when the caller proactively requested it by hitting the "O" key. Amazingly, 44.1% of the time, the caller hitting the "O" key is ignored. Hard to believe this – the caller specifically requests that they be connected to an operator and almost half of the systems tell the caller to "go pound sand". Another one – if the caller is obviously confused, in less than one third of the instances, does the automated system have the sense to immediately connect to the operator. For the other two-thirds of the time, these systems do other caller-friendly things like hanging up, telling the caller he/she made an error, repeating the prompt, etc. The Sterling Audits "Voice Response Usability Almanac, 2004" is priced at just $495. Sterling Auidits will provide the "Voice Response Usability Almanac, 2004" to ASRNews subscribers at a 10% discount. 877-504-5145 or 702-228-8556 www.sterlingaudits.com Gail Margulies gail@sterlingaudits.com

Telephone – IVR Applications

BCE Elix enables city of Montreal to become a technological pioneer Using innovative solutions to enhance citizen services BCE Elix (Montreal, Quebec) and the city of Montreal announced the launch of telephone applications that provide Montrealers with an enhanced 87-ACCES automated telephone service system. Innovative bilingual voice applications implemented by BCE Elix will now enable citizens to call the 87-ACCES telephone line at any time and quickly access information that was previously available only on the city of Montreal’s Web site, thus providing a flexible and simple access to municipal services ensuring superior quality of service 24/7 to all its citizens. Three interactive voice applications are now available, in French and English, for the city of Montreal: "Info-Collectes", an interactive application that provides schedules for garbage and bulky object collection, can now be accessed merely by having the citizen say their postal code. "Info-Remorquage" enables callers to locate their towed cars during snow-removal operations by saying their license plate number to the voice-enabled application. "Points of Service" enables Montrealers to obtain their borough’s points of service address, phone number and opening hours by merely saying the borough’s name. Given the complexity of some of the city of Montreal’s 27 borough names, this represented quite a challenge in English as well as in French. "The BCE Elix system will be complementary to the city of Montreal’s existing contact centre," said Martin Lefebvre, Head of Citizen Relations Management Division, city of Montreal. "Agents will still be available during office hours to accept 87-ACCES calls that require their help, but the BCE Elix applications will offer services 24/7 to our citizens, help us answer more calls, ease contact centre traffic, and allow us to be technological pioneers." Using the Web, speech recognition, text-to-speech, and VoiceXML and CCXML standards, this voice service application simplifies access to information by making online Web data available to phone users. Info-Collecte will be extremely handy for Montrealers moving during the peak season. In addition, Info-Remorquage will also be handy during snow-removal operations, when you don’t necessarily have Internet access, but have a mobile phone available. Contact: BCE Elix Rita Azrak, Director, Marketing and Communications Rita.azrak@bceelix.com Tel.: +1.514.768.1000, ext. 2323

Nestle Waters North America Selects Cisco Customer Voice Portal for Self-Service Voice Applications; Cisco Solution Redefines the Standard for Interactive Voice Response Systems Cisco Systems, Inc. (Nasdaq:CSCO) announced that Nestle Waters North America, Inc., is deploying the Cisco Customer Voice Portal (formerly Cisco Internet Service Node). The Cisco CVP's open platform allows enterprises to avoid "forklift" upgrades of existing contact center equipment and to protect their existing investments by using installed contact center technology. Cisco CVP and Cisco IP Contact Center (IPCC) solutions form the core of the Customer Interaction Network, allowing companies to extend customer service across an entire organization and transparently blend self- and assisted-service for an enhanced customer experience. The Cisco CVP supports both speech-enabled and touchtone applications that can be quickly integrated with backend data and business rules available through the Web, using J2EE and VoiceXML standards. The graphical development tools provided with the Cisco CVP help to enable assembly of the most complex voice applications faster and at lower cost than traditional IVR or other voice portal technology by using voice dialog components that facilitate reuse. The Cisco CVP uses a Cisco gateway to provide call termination and voice browser functionality. A Web server hosts the application. www.cisco.com. Contact: Cisco Systems, Inc., LeAnne Schrotzberger, 408-853-4336 (Press) lefrank@cisco.com Jane Courtenay, 408-526-7893 (Industry Analysts) jcourten@cisco.com Blair Christie, 408-525-4856 (Investor Relations) blchrist@cisco.com Kurt Mey, national technology manager for Nestle Waters North America, Inc.

Vancouver credit union upgrades call routing: The promise of advanced touch-tone banking capabilities and enhanced ease of use has prompted the Greater Vancouver Community Credit Union (GVC) to upgrade its telephone banking system. GVC was one of the 37 client sites throughout B.C. and Saskatchewan that upgraded to Datawest Solutions’ new Datawest Phone Banking system. Credit union members switching from Datawest’s legacy VisionLink phone banking system to the new solution will enable customers to: retrieve account balances with two keystrokes rather than the 20 or so keystrokes required under the older system; access other memberships they are associated with in a single call; arrange to be advised if a certain event occurs; and create new accounts through a telephony interface. "There are quite a few enhancements that they made that have been quite beneficial," said Trisha Tanginoa, manager of administration and marketing, GVC, which has about 6,500 members that are served from four locations. "We also have the ability to custom-route member calls, if they were experiencing difficulties, to particular branches if they so chose, which is beneficial to us because we don’t have a call centre. Telephone banking is an essential nowadays. We may not get a huge increase in penetration as far as our membership is concerned, but the people who use it are quite loyal to it, so it’s a product that we have to offer. And we’ve had some positive feedback from our membership. They like the new functionality. It was a really seamless implementation as far as our membership is concerned." According to Datawest, there are four keys areas to consider when evaluating the Datawest Phone Banking system: open architecture, flexibility from a configuration perspective, incorporation of technologies such as speech recognition, and user friendliness. "The product itself is able to communicate into multiple banking platforms," said Lesley Bernardin, implementation consultant, Datawest. "Therefore any financial institution regardless of the banking platform that they’re using is able to operate this telephone banking system." Because of the open standards approach taken by Datawest, clients won’t have to worry about being tied down to any one standard, stressed Iain Black, president of Datawest’s Banking Solutions Group. "In the banking system game, there’s extraordinary reliability that customers have on their core banking system provider," said Black, whose company was acquired recently by Open Solutions based in Glastonbury, Conn. Regulatory and shareholder approval are still pending. "It’s not unusual in historical banking systems to have a lot of your options limited by virtue of the decisions you made in years gone past in terms of the banking system and what it can talk to and what it can’t talk to." Bernardin said that Datawest will develop speech-related technologies for future product releases. "We are actually in the process of putting the final touches on the first enhancement under the speech recognition umbrella," she said. Another component the company is working on is "secure speech" which would identify the caller by voice. Contact: Datawest Solutions Inc. Shari Shamro Investor Relations Tel: 604.714.2421
ir@datawest.ca iain.black@datawest.com

Edify Voice and Speech Powers Top 4 Banks in China: Edify Corporation (Santa Clara, CA) announced that the top 4 banks in China all utilize Edify Voice and Speech to provide exceptional customer service, streamline business processes and reduce costs. Edify enhances the management of banking customer service operations for all of the top 4 Banks including Industrial and Commercial Bank of China , Agricultural Bank of China , China Construction Bank , and Bank of China as well as leading middle tier banks such as Everbright Bank. As a result, Edify enables them to increase call accuracy and accommodate higher call volumes while delivering more personalized and intelligent customer interactions. In particular, Everbright Bank, the financial subsidiary of The Everbright Group, recently implemented a distributed Edify voice self-service system that readily integrates customer self-service throughout its enterprise call centers. Everbrights Bank’s goal was to improve customer satisfaction while decreasing its long distance phone expenses between its more than 20 call centers throughout China’s major provinces. Everbright Bank looked to Edify to consolidate its service offering to provide a unified and efficient customer-facing solution. he Edify system enables Everbright Bank customers to simply use one number to reach a variety of self-service options, including banking, credit card and securities trading information. With the Edify enterprise voice self-service automatically directing 90 percent of the incoming calls, only 10 percent of the calls actually go through the bank’s internal network, streamlining business processes and reducing overall expenses. www.edify.com. http://www.s1.com. Edify Corporation Cynthia Dimidik, Edify Corporation: (408) 486-1791; cynthia.dimidik@edify.com said Wu Keyin, call center project manager at Everbright Bank

JOS, CTIL and InfoTalk Join Forces to Reengineer Customer Service Hotline for MTR Corporation Limited: Jardine OneSolution (JOS), Continuous Technologies International Limited (CTIL) and InfoTalk Corporation have joined forces to reengineer the tri-lingual Customer Service Hotline with a comprehensive call centre solution to upgrade the technology and increase the capability of the MTR Corporation Limited (MTRCL). Merging the ability and professional one-stop solution from JOS, extensive system integration experience from CTIL and cutting-edge speech technologies from InfoTalk, the new MTR customer service hotline is expanded to cope with the current and future business needs of the MTRCL. The MTRCL is continually increasing the capacity and coverage of their transportation network with six lines running over 87.7 km including a dedicated, premium Airport Express service. This fast growing transportation system services over 2.3 million people each weekday. This growth and high utilization rate has increased the demand on the older customer service hotline system to meet the customers' needs and expectations. The MTRCL is therefore looking for better technologies to completely revamp the system in light of providing better and more services for its customers. Based on the familiarity with MTRCL's infrastructure and understanding of their expectations, the three parties have delivered to the customer an elegant contact centre solution. The MTRCL has now added several new services to the customer service hotline, including voice-enabled Airport Shuttle Bus information enquiry and real-time marketing program information. With a current total of fifty (50) MTR and Airport Express stations, it was quite difficult for the older system to efficiently provide train fare and traveling time information through a touch-tone enquiry system because of the limitation in the number of keys. With enhanced speech technology in place, customers can now access fare and traveling time information simply by saying the entry and exit station names in Cantonese, Putonghua or English. It would be very cumbersome for customers to obtain Airport Express Hotel Shuttle Bus information with a touch-tone system because of the large number of hotels in Hong Kong. Now, travelers using the Airport Express only need to tell the system which hotel they want to go to and they can get the shuttle bus information right away. In addition to enhanced speech recognition, the new customer service hotline utilizes text-to-speech (TTS) technology, providing real-time information on ad-hoc train service adjustment and marketing promotion programs. As market information changes frequently, updating the system required tedious manual recording of the new information with the older technology. TTS technology provides an efficient and cost-effective way to ensure the latest train service and promotional information is available in the MTR customer service hotline. The state-of the-art TTS technology used in the MTR system provides a pleasant and natural sound which could easily be mistaken for a human voice. Other services provided in the customer service hotline include lost property information, fax information or information on facilities for passengers with a disability, left baggage service information, car park information and shareholder enquiry. The customer service hotline application was developed and integrated by Continuous Technologies International Limited (CTIL). CTIL worked closely with JOS and InfoTalk and leveraged its expertise in system integration of customer interaction technologies, helping the MTRCL build the customer service hotline based on its requirement. Jardine OneSolution (JOS) is one of the Asia's leading integrated IT service and solution provider and part of the Jardine Matheson Group. JOS has been a force in the IT industry for more than 50 years and today its customers include the best-known multinational and regional companies operating across Asia from industry sectors such as Financial Services, Trade and Commerce, Transportation and Logistics, Utilities and Telecom as well as Government. Today JOS delivers a rich blend of IT offerings ranging from Technical Services and Support and System Integration to Outsourcing and Managed Services. The company also leads the market in critical business application such as Procurement and Deployment Services - a capability that marries its depth of technological and logistical capability with rich experience in procurement and supplier management to create a compelling and increasingly popular offering. JOS employs more than 2,000 staff in eight offices in Hong Kong, China, Singapore and Malaysia. www.jos.com. Founded in 1989, Continuous Technologies International Limited (CTIL) is an independent systems integrator and one of Asia's largest call centre and customer interaction solutions providers. The Company is well known for its experience in the implementation of voice, call centre and telephony technologies. CTIL's customer base mainly consists of major banks, insurance companies, utilities, transportation and government agencies. The Company is headquartered in Hong Kong and has offices in Beijing, Shanghai, Nanjing and Guangzhou. Continuous Technologies embraces the principle of a customer-centric operation. For over 13 years, we have been helping businesses and organizations in China and Hong Kong to enhance their customer service capabilities. Our offerings include a suite of world class CRM and Customer Interaction Management (CIM) solutions. As a turnkey systems integrator, we have the experience and knowledge to implement CRM, Call Centre and Interactive Voice Response (IVR) systems in a multi-vendor environment often involving integration with legacy systems and applications. www.ctint.com. Contact: info@infotalkcorp.com or +852 2190 9600 Dr. Alex Leung, CEO of InfoTalk. Continuous Technologies, Telephone: (852) 2850 6338, Antonio Cheung, General Manager of CTIL, JOS, Lau Siu Pong, 852-2880-3090, splau@jos.com.uk Mr. Thomas Wan, Group Director of JOS, Mr. Daniel Lai, Head of Information Technology of MTRCL.

Countrywide Chooses Solutions From NMS Communications and Nuance to Enhance Customer Experience: NMS Communications (NASDAQ: NMSS) announced that Countrywide Financial, a leading diversified financial services provider, has chosen the NMS Open Access platform and Nuance's (NYSE: NUAN) speech recognition technology, for its call center applications. The combined NMS/Nuance solution is designed to provide speech recognition capabilities, and a path to add IP capabilities, drive efficiencies, and enhance the customer experience. NMS' Open Access platform (www.nmscommunications.com/OpenAccess) supports critical call center functions and speech-driven applications on both TDM and IP networks, giving customers a solution that can evolve along with their networks. The NMS solution is also designed to allow other applications to be incorporated, with minimal network disruption. Founded in 1969, Countrywide Financial Corporation is a member of the S&P 500 and Fortune 500. Through its family of companies, Countrywide provides mortgage banking and diversified financial services in domestic and international markets. Mortgage banking businesses include loan production and servicing principally through Countrywide Home Loans, Inc., which originates, purchases, securitizes, sells, and services primarily prime-quality loans. Also included in Countrywide's mortgage banking segment is the LandSafe group of companies that provide loan closing services. Diversified financial services encompass capital markets, banking, insurance, and global, largely through the activities of Countrywide Capital Markets, a mortgage-related investment banker; Countrywide Bank, a division of Treasury Bank, N.A., a banking entity offering customers CDs, money market accounts, and home loan products; Balboa Life & Casualty Group, whose companies are national providers of property, liability, and life insurance; Balboa Reinsurance, a captive mortgage reinsurance company; and Global Home Loans, a European mortgage banking joint venture in which Countrywide holds a majority interest. www.Countrywide.com. Contact: NMS Communications, Pam Kukla, 508-271-1611 Pam_Kukla@nmss.com Lior Ofir, VP, Enterprise Voice Technologies for Countrywide

Bell Canada adopts Vicorp xMP: Vicorp announces that Bell Canada will be deploying the Vicorp xMP Application Development Platform within its Next Generation IP network. The Vicorp VoiceXML platform in conjunction with the VoiceGenie media server will enable Bell Canada to add powerful new advanced speech interface options to its successful portfolio of calling card services for Enterprise and Residential customers. Designed to capitalise on open standards such as VoiceXML and the J2EE environment, Vicorp’s xMP solution tool suite enables the creation, management and delivery of services with intuitive bi-lingual speech interfaces, messaging, content delivery and user personalisation. The Vicorp xMP platform hides the complexities of application development - enabling service providers to deliver exciting new applications, rapidly and cost-effectively. Bell Canada is Canada's largest communications company and provides local telephone, long distance, wireless communications, Internet access, data, satellite television and other services to residential and business customers through some 26 million customer connections. www.vicorp.com Contact: Scott Sibuns, VP Worldwide Sales for Vicorp Tel: +44 (0) 1753 660 500, sales@vicorp.com

Continental Airlines Elevates Customer Service with Voxify’s Automated Agents: Voxify (Alameda, CA) announced that Continental Airlines (NYSE: CAL) is successfully using Voxify’s Automated Agents to manage flight reconfirmation calls and deliver excellent customer service. Continental chose Voxify for their breakthrough technology and ability to deliver automated systems with superior customer experience, rapidly and cost-effectively. Continental estimates that approximately 10% of incoming customer calls involve flight reconfirmation. With Voxify’s Automated Agents successfully handling these calls, Continental’s live agents can take thousands of calls every day for sales and other customer services. Automated Agents will also help in reducing overall hold times and increasing Continental’s customer service capacity.  Customer satisfaction with Voxify's Automated Agents is already high, with over 70% of customers using the system to complete their reconfirmation. "Not only have we seen high completion rates within the first few weeks of deployment, Continental has already achieved a 100% ROI in the three months," said Martin Hand, Staff Vice President Reservations and City Ticket Offices at Continental. "Continental's call centers handle over 40 million calls a year. With Voxify, we have been able to deliver a high quality of service, while realizing significant savings through automation. Voxify is an important part of our plan to deliver a consistent and excellent customer experience." Voxify Automated Agents are built for advanced customer service calls, such as fare finder and reservations, and can accommodate customers with varying amounts of information. For example, Voxify’s Reconfirmation Automated Agents respond to customer inquires about their existing reservations and provide them an ability to reconfirm their flights, regardless of how much information the traveler has on hand.  High conversational usability requires complex call path maps to handle the many variant call paths that a customer service call can take.  Voxify has solved this challenge with its patented Conversation Engine technology, a breakthrough for automating the development and maintenance of highly usable speech applications.   The Conversation Engine imbues the Automated Agents with advanced conversational skills, enabling them to engage callers in sophisticated dialogue, and understanding multiple pieces of information in a single response or adjusting to unexpected information provided by the caller. As a result, callers experience an Automated Agent who understands how to give them what they need quickly and easily, closely approximating the effectiveness of a human being. Voxify’s Conversation Engine contains built-in performance tracking and detailed analytics on call paths, including decision-making, drop-off points, and likelihood ratios, helping Automated Agents continuously improve call handling and conversion rates. Voxify’s real-time reporting provides Continental immediate enhanced visibility and understanding in this customer service area, allowing Continental to leverage this business intelligence for ongoing training and system development purposes. "The history of the travel industry shows that the best customer experiences come from agents who deliver personalized service. Customers expect the personal touch in every interaction -- even if the voice on the other end isn't a human being. We’ve endeavored to this standard by delivering a personal touch with every customer interaction. By using Voxify, we’re taking a giant leap forward in this, which will pay significant dividends for our valued customers," said Continental’s Hand. Powered by the company's patented Conversation Engine, Voxify Automated Agents model the intelligence and personality of a trained live agent, allowing them to engage callers in sophisticated dialogue to perform advanced customer service functions, like reservations, sales, and account requests. Voxify Automated Agents are helping call centers deliver new services, close more sales, and handle unplanned call volumes without hiring additional staff. Voxify was founded by MIT technologists and enterprise software veterans. www.voxify.com. Contact: Hollis Chin, Voxify 510.545.5015 hchin@voxify.com Martin Hand, Staff Vice President Reservations and City Ticket Offices at Continental

Voxify Automated Agents Help Wyndham Roll Out a Virtual Red Carpet: A team of Automated Agents has been hired by Wyndham International to give its guests a red carpet greeting and white glove treatment any time they call for service. The Agents are provided by Voxify, the company to create Automated Agents for call centers with the conversational skills to handle advanced customer service calls, and will greet guests, help locate hotels, book reservations, or update their loyalty program information. Wyndham chose Voxify because its patent-pending technology that models the intelligence and service demeanor of human agents, has seen wide acceptance in the travel and hospitality industries. Plus, as a hosted solution, it can be customized and deployed within weeks rather than the months usually required for a custom-developed system. Wyndham operates more than 145 upscale hotels and resorts offering more than 36,000 rooms globally. More than 2 million prospective guests at these hotels reach Wyndham's call center annually through its toll free reservation numbers, making it one of the hotelier's most critical service points. Voxify Automated Agents will enhance the experience by providing a smart, central system that will greet guests immediately, identify what they need, and instantly route them to the best available resource, whether human or automated. For example, Automated Agents might be best suited for rapidly reconfirming reservations. On the other hand, live agents would be best equipped to help a guest with an unusual question about a particular hotel property. Initial plans for the Voxify-powered call center projects 25 to 40 percent of service calls to be fully automated. Indeed, roughly one million of Wyndham's calls for reservations, confirmations, or loyalty program requests will be handled from greeting to completion by an Automated Agent. But in keeping with Wyndham's service pledge, callers will always be given the option to reach a live agent for any request. The Automated Agents will serve as the vehicle for connecting each caller with the right resource in the least amount of time. "We believe in giving our guests choices on how they want to be served," stated David Mussa, Wyndham’s Vice President of Reservations. "A large percentage of the calls we receive at our call center are common requests that can be easily and efficiently fulfilled by an intelligent Automated Agent. Offering our customers additional options in facilitating their requests, allows us to deliver a better guest experience – which is at the core of Wyndham’s brand philosophy." Voxify Automated Agents are built for advanced customer service calls, such as fare finder and reservations, and can accommodate customers with varying amounts of information. High conversational usability requires complex call path maps to handle the many variant call paths that a customer service call can take.  Voxify has solved this challenge with its patented Conversation Engine technology. The Conversation Engine imbues the Automated Agents with advanced conversational skills, enabling them to engage callers in sophisticated dialogue, and understanding multiple pieces of information in a single response or adjusting to unexpected information provided by the caller. As a result, callers experience an Automated Agent who understands how to give them what they need quickly and easily, closely approximating the effectiveness of a human being. Voxify has deep experience in the travel industry, serving some of the industry's largest and most successful airlines and hotel chains. Among its other clients are Continental Airlines, World Choice Travel, part of Travelocity, and Pegasus Solutions. Voxify is the first company to create automated call center agents with the conversational skills to handle advanced customer service calls. Powered by the company's patented Conversation Engine, Voxify Automated Agents model the intelligence and personality of a trained live agent, allowing them to engage callers in sophisticated dialogue to perform advanced customer service functions, like reservations, sales, and account requests. Voxify Automated Agents are helping call centers deliver new services, close more sales, and handle unplanned call volumes without hiring additional staff. www.voxify.com. Contact: Hollis Chin, Voxify 510.545.5015 hchin@voxify.com

Telephone - IVR Products/Services

Apex Voice Communications Announces OmniVox for Windows XP: APEX Voice Communications (Sherman Oaks, CA) announced the general availability of OmniVox version 5.0, an Intelligent Call/Media Processor supporting Windows XP. Some of the new features of OmniVox for Windows XP include the Call Recording Module, the Karaoke Module and support for more ScanSoft products. Additionally, a number of key enhancements have been made to the OmniView Service Creation Environment, the award-winning graphical design, development, deployment and maintenance module. With the new Call Record Module, OmniVox for Windows XP records activities from both the IVR and the telephone. From the IVR, OmniVox records the prompts which are played to the caller, while from the telephone, OmniVox records both the spoken responses of the caller and the DTMF keystrokes entered by the caller. The Call Record feature is used widely for transaction-based call processing, such as banking, stock trades and purchases (order entry), where the service provider requires a complete activity log of the call (i.e. in cases of disputes). The Karaoke Module plays a background audio file and then simultaneously records what the caller is singing (or speaking). Already installed across the Middle East, this Module was originally marketed as a way to send Karaoke songs to friends and family. However, over the last few years, the entertainment industry around the world has begun to use this type of module to increase home audience participation, primarily in talent shows and game shows. In the United States, this type of Karaoke Module is used in the popular television talent show American Idol. OmniVox has supported automated speech recognition technology for over 10 years, as part of its standard offering. ScanSoft’s OpenSpeech is a high-performance speech engine that is open, standards-based, and tuned to VoiceXML, delivering enhanced levels of accuracy specially in wireless environments. With the addition of ScanSoft’s OpenSpeech Voice Recognition, OmniVox now supports automated speech recognition from both ScanSoft and Nuance, two of the leading companies in the field. Much like voice recognition, OmniVox has supported text-to-speech for a number of years. With the addition of ScanSoft’s Speechify (SpeechWorks Division), OmniVox now supports two text-to-speech technologies from ScanSoft, giving greater options to developers worldwide. Text-to-speech is used for a wide range of applications, including delivering emails for unified messaging systems, providing prompts for IVR applications, as well as speech-enabling voice portals, directory assistance and auto-attendant applications. ther new features of OmniVox for Windows XP include a utility to facilitate the porting of OmniVox applications from UNIX or Linux to Windows; the ability for the SQL Commands to handle Binary Large Object (BLOBs), such as voice files, pictures and spreadsheets; and expanded support for SS7, which now includes Datakinetics from Intel.APEX is a leading manufacturer of enhanced services and real-time billing solutions. APEX’s high-density, open platforms enable innovative, revenue-generating services to reach the market quickly, easily and cost-effectively. APEX products are engineered for both existing and next-generation networks, offering circuit-based providers a migration path to packet-based services. www.apexvoice.com Value Added Resellers may also visit the APEX Reseller Program at www.arp.apexvoice.com Tel: +1 818.379.8400, Elhum Vahdat, executive vice president of APEX Voice Communications, sales@apexvoice.com

Audentify Launches Intelligent IVR System: Audentify, a leading provider of next-generation contact center technology and a division of Autonomy Corporation plc (Nasdaq: AUTN; LSE: AU.), announced the launch of Audentify Intelligent IVR, an Interactive Voice Response system that allows organizations to effectively automate customer service operations in the contact center by truly considering customers' needs. Intelligent IVR forms an accurate understanding of a caller's query in real-time and automatically acts upon this information, providing an intelligent and dynamic response. Facing constantly high call-volumes in their contact centers, more and more enterprises are seeking ways to automate their customer services in order to answer queries faster and more efficiently. However, pre-set options often do not match callers' needs and lead to frustration when the system cannot provide a sufficient response. Unlike many other solutions, Audentify's Intelligent IVR allows organizations to form an accurate understanding of a customer's query and immediately act upon this request. The system permits callers to explain a particular problem in their own words, understands the conceptual meaning of this query and automatically returns useful results. If a relevant answer is not found within the system, Intelligent IVR instantly routes the call to a qualified agent who will be able to give additional advice. In this way, the system ensures that calls are handled as efficiently as possible, easing the burden on agents and enabling them to focus on those calls that need their immediate attention. At the same time, Intelligent IVR allows customers to benefit from a faster and more accurate service, eliminating frustration and the feeling of facing a rigid and impersonal system. Built on a combination of Autonomy's ability to form a conceptual understanding of any information and patented voice phoneme level pattern matching, Audentify's Intelligent IVR system has a wide-vocabulary speech recognition capability and is completely speaker and language independent. Audentify, a division of Autonomy Corporation plc (Nasdaq: AUTN; LSE: AU.), is a provider of next-generation contact center technology. Built on Autonomy's Intelligent Data Operating Layer (IDOL), Audentify claims to be the first contact center technology able to understand the content and context of customer interactions such as email and telephone conversations in real-time and dynamically deliver pertinent content to users. Audentify enables contact centers to reduce call handling times, increase the quality of customer interactions, monitor individual agent performance, analyze call handling trends and enable all calls and interactions to be recorded and made instantly searchable. Autonomy Corporation plc (Nasdaq: AUTN; LSE: AU.) is a supplier of infrastructure software for the enterprise. Autonomy's technology powers applications dependent upon unstructured information including call center, customer relationship management, knowledge management, enterprise portals, enterprise resource planning, online publishing and security applications. Autonomy's customer base includes more than 1,000 global companies including BAE Systems, Ford, Ericsson, Royal Sun Alliance, Sun Microsystems and public sector agencies including the U.S. Department of Defense, NASA and the U.S. Department of Energy. Strategic reseller and OEM partners include ATG, BEA, Business Objects, Citrix, Computer Associates, EDS, IBM Global Services, Novell, Novient, Veritas, Vignette, Supportsoft and Sybase. www.audentify.com Contact: Beth Ladd, Autonomy Corporation plc +1 415/243-9955 bethl@us.autonomy.com Charlotte Herbert, +44 (0) 1223 448 000 charlotteh@autonomy.com

Avaya Announces New Communications Applications that Utilize Automated Speech to Drive Business Productivity and Enhance Customer Service: Avaya Inc. (NYSE:AV) announced new releases of its unified communication and contact center solutions, which deliver self-service functionalities across an organization through speech-enabled applications. Avaya Unified Communication Center and Avaya Interactive Response are core solutions within the company's portfolio of Internet protocol-based MultiVantage Communications Applications — which encompasses telephony, contact centers, messaging and unified communication. With the introduction of Avaya Unified Communication Center 2.0, Avaya significantly increases the scalability of unified communications, supporting up to 10,000 subscribers. Expanding upon user support helps large and growing organizations meet the needs of an increasingly mobile global workforce. Other new enhancements include integration with IBM Lotus Notes calendar and contacts, allowing users to utilize speech with this application to access schedules and contacts (in addition to pre-existing e-mail speech access). A single, easy-to-use administrative interface enables easier management through centralized administration. Avaya Interactive Response 1.2 adds greater international support, expanding its speech recognition and text-to-speech capabilities to more than thirty languages, including Japanese, Chinese, Korean and Italian. It also operates on higher-capacity servers, providing greater port density and more processing power for self-service applications. With the new Avaya Speech Applications Builder, businesses have all of the tools and components required to create an open standards-based environment that accelerates the cost-effective design of speech automation services. Designed for the Avaya Interactive Response platform, the Speech Applications Builder incorporates the VoiceXML standard and contains an expansive library of more than 100 pre-built functionalities (i.e. user designs, prompts, retry logic, etc.). This helps reduce the time and costs associated with the development, testing and maintenance of speech applications. In a preliminary installation using the Avaya Speech Applications Builder, one financial services company was able to automate three services — transferring funds, checking balances and verification of identity — within half the typical deployment schedule required for development of a single application using traditional deployment methods. * Avaya Unified Communication Center 2.0: www1.avaya.com/enterprise/uc/ucc/unified_communication_center/index.html *Avaya Interactive Response 1.2: www1.avaya.com/enterprise/brochures/gcc1992.pdf *Avaya Speech Applications Builder: www1.avaya.com/enterprise/cc/ss/speech/index.html Avaya Inc. designs, builds and manages communications networks for more than one million businesses worldwide, including over 90 percent of the FORTUNE 500. www.avaya.com contact: Jonathan Varman, Avaya 908-953-6432 jvarman@avaya.com

Cisco Targets CRM Integration: Cisco Systems is giving contact centers an early peek at what's under the Christmas tree. The call center giant announced today, about two months prior to its release, that version 3.0 of the Cisco Customer Voice Portal (CVP), formerly Cisco Internet Service Node, will be available in December. The CVP supports speech-enabled and touch-tone applications that can be integrated with back-end data and business rules accessible through the Web, using Java 2 Enterprise Edition and VoiceXML standards. One of the new version's major elements is that it allows customers to leverage their existing contact center investments. Companies can also integrate speech recognition tools to the CVP. In addition, the CVP provides a voice XML based IVR and the ability to create a unified virtual contact center using an IP-based network to integrate the technologies, including disparate ACDs. As a result, customers should have the flexibility to migrate to IP as their business needs allow. The company also professes that integrating TDM/IP contact centers with the IP switching capability of the Customer Voice Portal will result in slicing telecom charges through toll bypass between locations, and eliminating take back and transfer charges, carrier announcement charges, and ICM dip charges. Additional features include queuing calls at the edge of the network and the ability to integrate existing IVR, ACD, and CTI agents. www.cisco.com Contact: Steve McElderry, product line manager for Cisco's contact center business unit. Cisco Systems, Inc., LeAnne Schrotzberger, 408-853-4336 (Press) lefrank@cisco.com Jane Courtenay, 408-526-7893 (Industry Analysts) jcourten@cisco.com Blair Christie, 408-525-4856 (Investor Relations) blchrist@cisco.com

New Hosting Solution Helps Partners Meet Demands of the Rapidly Expanding Speech Industry: CPT International announced the immediate availability of its Voice Harbor outsourced hosting solution for voice application developers, technology providers and their customers. CPT is currently forming partnerships with voice application developers and technology providers to offer customers a faster, more cost-effective and feature-rich approach to launching voice application projects than the traditional customer premise build-out. Customers maintain all the control of an on-premise solution, but leverage the strength of a road-tested platform that operates best-of-breed voice technologies in world class, secure facilities. Application development partners value CPT because the capital investment hurdle to deploying speech applications is significantly reduced by hosting the application with Voice Harbor. Instead of investing time and money in building an on-premise telephony platform, their customers can immediately engage in defining and executing on the business value of their voice application project, shortening the overall deployment cycle. Technology partners benefit by gaining value from Voice Harbor's ability to aggregate and capture customers who may have foregone potential telephony projects due to the risk of up-front investment. Additionally, Voice Harbor serves these partners as a showcase environment for exposing new features and innovations to a community of voice application developers and enterprise customers. Voice Harbor's scalable, reliable, flexible hosting solution completely eliminates the telecom headache for developers and their clients. Its established, robust infrastructure offers flexible call capacity, predictable costs via usage-based pricing, and rapid deployment. Built on open standards, including VoiceXML and SIP, Voice Harbor supports speech recognition, text-to-speech, touch tone and call-center processing applications, in addition to advanced services such as reverse number lookup and speaker verification. With more than a decade of experience in delivering robust telephony performance, CPT meets the stringent security criteria and capacity needs of telecom carriers, government agencies and business enterprises. The company's facilities offer 24X7 security and climate control, continuous power and cooling systems, hardened architecture, and secure VPN connections. CPT International Inc. is a telecom services provider located in Marietta, GA. Our flagship offering, Voice Harbor, provides hosting services for VoiceXML, speech, multi-modal and traditional telephony applications. Voice Harbor brings flexible call capacity, predictable costs and rapid deployment to customers and partners seeking to outsource voice application infrastructure. With more than a decade of experience delivering robust telephony performance and proactive support to carrier-grade customers, CPT is the partner of choice for standards-based voice application hosting. www.cptii.com Contact: David Cohen, Voice Harbor Product Manager, 770-952-5399 marketing@cptii.com

Edify Introduces Licensed Speech Application For Banking Institutions: Edify Corporation (Santa Clara, CA) introduced Edify Voice Banking, a fully licensed, software application that enables banking establishments to automate customer interactions and provide access to account, branch, and other financial information through speech-enabled self-service. Based on industry standards such as VoiceXML and SALT, the Edify Voice Banking application eliminates enterprise customer concerns over application functionality and compatibility issues. The Edify Voice Banking product includes key features such as account information and history, stop pay, change PIN, funds transfer, as well as business enterprise features, such as, detail reporting and monitoring for easy, ongoing technical support and upgrades. With Edify Voice Banking, customers calling into the system can verbally identify themselves, their account, and their requests without having to traverse a complex touch-tone menu. Customers benefit from being able to interact with a friendly, conversational system that provides efficient access to the available banking information they might need. Edify Voice Banking brings together a proven software platform, application support, and voice services in conjunction with knowledgeable industry experts. As a result, Edify Voice Banking addresses the challenges confronted by financial institutions that want to, not only retain their customers, but ensure that up sell activities regularly take place. By offering enhanced customer interactivity with back-end compatibility, banks can now look to on-going, profitable relationships with their consumers. Some examples of the types of modules in the Edify Voice Banking application include: • Account Information • Loans • Certificates of Deposit and IRA • Funds Transfer• Merchant Check Verification• PIN Change • Banking Hours • Check Reorder • Lost or Stolen Credit Card • Instructions Menu • Call Transfer • Transaction Fax. Applications built with Edify are scalable, multilingual and flexible, allowing companies to easily integrate backend systems with multiple contact interfaces. www.edify.com. Inquiries Edify Corporation Cynthia Dimidik, Edify Corporation: (408) 486-1791; cynthia.dimidik@edify.com

MTI Offers Unique Mix with Speech IVR and Live Agent Services Under One Roof: MTI announced the expansion of its live agent service offerings to deliver greater flexibility to its speech IVR hosting customers. In addition to providing highly-trained professional agents and advanced notification capabilities, the company now offers Spanish language agent resources, web-enabled dual flat-screen agent stations, and in-network call patching. MTI’s service expansion delivers a broader set of options to companies wanting live agent backup to their speech applications. Now customers needing Spanish agent resources can turn to MTI for bilingual live agents. The company’s implementation of dual flat-screen agent stations allows agents to simultaneously interact with the caller according to a script and access customer websites or intranets for data entry, level 1 customer support, scheduling, and a wide variety of other applications. MTI’s new in-network call patching service allows the agent to direct the call back to the speech application or any other number. MTI offers a wide range of live agent services as well as its speech IVR services. This enables the company to offer turnkey solutions to companies needing high-quality opt-out live agent services working in tandem with their speech applications. While many companies are taking advantage of the advances in speech recognition technology to lower customer service costs, there is still often a need for people to speak with a live person. MTI (Message Technologies, Inc.) is an industry leader in providing Interactive Voice Response (IVR) and speech technologies. Its mission is to deliver innovative and affordable interactive voice solutions with maximum customer satisfaction. With over 2000 customers throughout North America and Canada, MTI specializes in speech IVR outsourcing services, including VoiceXML and SALT hosting. MTI also provides senior telephony and speech professional services for the design, development, deployment and ongoing testing of speech-enabled applications. It handles every part of the speech application infrastructure including trunk and platform provisioning, business continuation planning, monitoring, web access, physical and logical security and co-location, and live agent services. www.messagetech.com Contact: Lynn Abramson MTI 770-240-7410 lynn.abramson@messagetech.com Mark Abramson, CEO of MTI

S1 Enterprise Voice Powered by Edify Enhanced Speech-based customer Interactions: S1 Corporation (Nasdaq: SONE), a leading provider of integrated front-office applications for financial institutions, and its subsidiary Edify Corporation introduced S1 Enterprise Voice (powered by Edify). The product is part of the recent release of S1 Enterprise 3.0 – a truly integrated multi-channel solution that addresses the entire front-office of banks and insurers. With nine major applications - including voice, branch, call center and Internet solutions - on a common platform, S1 Enterprise 3.0 uniquely helps institutions lower operating costs, increase revenue opportunities, and improve customer service and retention by integrating and communicating customer information, transactions, and behaviors across channels. With Edify's proven voice and speech applications integrated on the S1 Enterprise Platform, global financial institutions can rapidly deploy leading voice capabilities and optimize the benefits of multi-channel integration. With S1 Enterprise Voice (powered by Edify), S1 customers can now benefit from: * Scalability - With voice and speech deployments from four ports to more than 8,000 ports, financial institutions receive the most flexible and reliable solution for building and deploying highly scalable enterprise voice self-service applications. * Robust voice and speech functionality - With leading-edge technologies like speech recognition, speaker verification, and text-to-speech, Edify's proven technology completely automates transactions formerly requiring live agent assistance. * Increased operational efficiency - Financial institutions can service large volumes of customer inquiries through automation, offering customers more timely resolution and freeing up live agents for more complex duties, all while driving down overall costs.* Enhanced security authentication - Voice solutions using speaker verification offer greater security for authenticating users than personal identification numbers. * Application compatibility - Support for VoIP and standards such as VoiceXML and SALT eliminates concern over application functionality and compatibility issues allowing institutions to maintain their Edify and/or standards VoiceXML applications on any infrastructure - VoIP or standard public switched telephone network (PSTN). www.s1.com www.edify.com Edify Corporation Cynthia Dimidik, Edify Corporation: (408) 486-1791; cynthia.dimidik@edify.com

Self-Service Usability Almanacs Debut: Sterling Audits (Las Vegas, NV) announced the availability of two new landmark research reports entitled: Web Site Usability Almanac 2004 and Voice Response Usability Almanac 2004. The publications are the result of a six-month study on self-service usability based on the Sterling Audits Usability Index, a standard research index for service automation systems. A major trend in enterprise communications is the consolidation of all customer-facing activities into a more cohesive whole. The catalyst for this consolidation deals with a growing demand on behalf of customers to have more seamless access to customer service across all touch points (i.e. Contact Centers, Web Sites and Automated Telephone Systems). Until recently, there was no standard or published methodology for scoring and benchmarking the overall usability of these systems. The Sterling Audits Usability Index was developed to provide customer service executives with a tool to implement their own research instrument for scoring and weighting the effectiveness of self-service systems. Over 220 specific indices are measured – each rolling-up to top metrics including Navigation; Content; Usability; Interactivity; and Credibility. Scores for these critical areas are tabulated using web-based software called the Sterling Audits Research Portal. The details of the Index, including instructions on how to implement your own research instrument is available from Sterling Audits. The research corpus is based on the output of a joint research initiative sponsored by Sterling Audits' own research staff and its research affiliates. These research affiliates include BenchmarkPortal, Elrod Interactive, Intelemedia, OPUS Research, and WesBro associates. During the past six months, over 100 researchers in North America were involved in the collection of over 45,000 data points. Each researcher used the Sterling Audits Research Portal to log observations and answer survey questions based on Web Sites and Voice Response Systems used in the course of everyday business by each researcher. This provided access to systems requiring user identification and passwords. Issues covered include: Information on transactional integrity, persona, menu strategy, organization of content and interface design. The research is ongoing, so updates will be made in annual editions of the Almanacs. Both the Web Site Usability Almanac 2004 and the Voice Response Usability Almanac 2004 cover nine industry segments including: Communications; Financial Services; Government; Healthcare; Insurance; Manufacturing; Retail; Transportation; and Utilities. Each Almanac is arranged by chapter. There is an introductory chapter, and a chapter on approach and methodology. The remaining chapters are dedicated to the survey results based on the aforementioned top metrics. A final chapter is dedicated to industry segments reveals the upper quartile, median and lower quartile averages for each critical area. There are over 200 graphs and charts in the Web Site Almanac and over 100 in the Voice Response Almanac. Two other publications: How to Audit and Benchmark Your Web Site and How to Audit and Benchmark Your Voice Response System, are also available. These books show readers how to score and weight their own self-service systems so they can be benchmarked against industry peer data from the Almanacs. In addition, a two-day workshop is held quarterly for certification in the use of the Sterling Audits Usability Index. The "how to" books are the basis for classroom tutorials. Each Almanac sells for $495.00. To order, you can call Sterling Audits’ toll free number (877-504-5145) or visit www.sterlingaudits.com to download the book order form to fax- or mail-in orders. Corporate checks, money orders, American Express, VISA and MasterCard are all acceptable forms of payment. Shipping options include U.S. Mail and FedEx. Sterling Audits provides usability testing and benchmarking services for all self-service operations — Web Sites, Voice Response and automated parts of the Contact Center. The company offers consulting, research reports, books and educational workshops on their standard usability benchmarking methodology called the Sterling Audits Usability Index. The company helps its clients to reduce the cost of customer service through incremental improvements in self-service automation. Part of the Sterling Audits vision is the hosting of a sophisticated research portal that allows analysts and end-user customers to easily log on to score self-service systems. Based on this research, the company publishes educational materials and teaches workshops in the spirit of knowledge transfer. Toll-free 877-504-5145 or outside the U.S. 702-228-8556. www.sterlingaudits.com Gail Margulies Sterling Audits 702-228-8556 gail@sterlingaudits.com

Syntellect Announces Vista 5.0: Syntellect Inc., announced that it will release the next major version of its flagship product Vista in the fourth quarter of 2004. Vista, the industry's first open standards, software-centric IVR platform was originally introduced to the market in 1998. Since then customers have leveraged Vista's next generation IVR architecture to deploy touchtone, advanced speech recognition, and text-to-speech solutions in both VXML and Java, resulting in thousands of deployed applications around the world. Vista 5.0 will integrate Nuance 8.5 and Verifier 3.5, as well as Vocalizer, to offer Syntellect customers the latest advancements in speech recognition and text-to-speech technology. With added support for Windows Server 2003 and VoIP, supported by an existing suite of development tools for quickly and easily creating high-quality speech applications, Vista 5.0 affords Syntellect customers the opportunity to implement state-of-the-art self-service with a fast return on investment and a low total cost of ownership. Interchange, Syntellect's new contact center automation solution, will also take advantage of the new features of Vista 5.0. Customers can now deploy a fully integrated contact center solution to help them automate, optimize, and virtualize their customer interactions. 1.800.788.9733 or in the UK on +44 (0) 1628 897 500 and say "Syntellect," www.Syntellect.com/. Contacts: Diana M. Ferguson, Manager, Marketing dferguson@Syntellect.com Tel: 602 789 2884

TuVox Unveils Speech-Enabled Enterprise Call Routing App: Speech application developer TuVox has launched a new speech-enabled call router, TuVox Perfect Router. The software uses a conversational approach to route calls to appropriate destinations. The router can use the TuVox Perfect Connect agent transfer application to pass information about callers to agents' desktops, eliminating the need for callers to repeat information. TuVox Perfect Router is built on TuVox's patent-pending SmartGen approach that converts existing enterprise content, such as web self-service or knowledge base content, into speech applications. TuVox says that, unlike other call routing solutions available in the market, TuVox Perfect Router can be deployed in weeks rather than months. The application is platform independent, enabling deployment on any voiceXML or speech application language tags (SALT) platform and with any speech recognition engine. The software features database-driven routing, which allows companies to update routing destinations without re-building their applications. TuVox Perfect Router also ships with TuVox CVR 4.0 application lifecycle management tools and components, which help enterprise users maintain and extend their applications. TuVox provides enterprise software to revolutionize voice self-service. TuVox creates a superior caller experience through sophisticated speech applications that are fast to develop and easy to deploy and maintain. Built completely on open standards, TuVox's enterprise software can automate virtually any type of call - including natural language call routing, self-service transactions, information requests and product support. TuVox has pioneered industry-leading SmartGen technology that allows enterprises to quickly create highly sophisticated and conversational speech applications from their existing content, data and business processes. TuVox's mission is to create voice automation that is so effective that callers will freely choose it over a live agent, and to make voice automation so cost-effective that any enterprise can deploy it. TuVox customers include companies such as Activision, CBCA, Columbia House, Definity Health, MCI, Netflix, Oriental Trading Company, Roxio, TiVo, and the United States Postal Service. www.tuvox.com

Unveil Technologies Announces Version 2.0 of its Conversation Manager Voice Self Service Application: Unveil technologies, Inc., (Waltham, MA) announced version 2.0 of its Unveil Conversation Manager. Unveil Conversation Manager is voice application software that enables call centers to deploy software-based virtual agents that automate, assist, analyze and route customer service calls. The new version of Unveil’s software includes enhancements to Adaptive Learning aimed at significantly reducing the cost of ownership for speech, integration with third party systems from key partners, and robust system statistics and reporting. These and other key improvements build upon the product’s features aimed at reducing call center operating costs while increasing caller satisfaction. The new capabilities of the Conversation Manager software address key demands of the call center marketplace – integration with existing systems, reporting on key metrics, and reducing the total cost of ownership for speech applications. This version of the software integrates with additional critical telecommunications systems from key partners and provides detailed reporting on system automation rates and caller behavior. It also strengthens the patented adaptive learning techniques which are aimed at dramatically reducing speech tuning and maintenance costs. "As an outsourcer of customer service, we are measured by our ability to deliver a consistent high quality of care. Unveil’s software helps us maintain high customer satisfaction rates while also increasing speech automation and agent productivity," said Susan Cohen, Vice President of Marketing and Business Development for UpSource, Inc. "We expect the expanded reporting and learning capabilities in the latest release of the Conversation Manager will enable us to pinpoint opportunities for further automation, and help us cost-effectively achieve higher automation rates over time." Unveil Conversation Manager helps call centers balance service quality and costs, so that they can increase customer satisfaction while reducing operating expenses. Using Conversation Manager, call center managers can deploy virtual agents that replicate their live agents. Unveil virtual agents engage callers in flexible automated conversations, collaborate with live agents to increase labor efficiency, and adapt to callers to get better over time. Virtual agents work with customer service representatives by pre-collecting and routing information or by automating entire customer dialogs. Unveil’s application software lets call centers optimize each interaction by applying the right mix of resources – speech automation, live agent, or both – to both achieve customer satisfaction and cost reduction. The product’s key features include: • Conversation Assist, which enables Unveil virtual agents to get help from live agents during a customer dialog, extending self service automation and increasing labor efficiency. • Adaptive Learning, an automated process that analyzes customer conversations to dynamically improve system performance over time. • Meaning Recognition Engine, which lets the caller engage virtual agents in unscripted, conversational speech, increasing end user comfort and satisfaction. Version 2.0 Key Feature Enhancements Learning Server – The Learning Server powers Conversation Manager’s Adaptive Learning capabilities. The Learning Server logs and organizes information captured from customer service interactions handled by live agents, improving system performance over time. Key improvements to the Learning Server enable call center business analysts to review, approve and add thousands of new caller utterances to the speech grammars and underlying meaning recognition engine, dramatically reducing ongoing speech tuning costs. The adaptive learning process does not require linguistic skills or speech recognition expertise, and can be managed by a call center business analyst via an intuitive software wizard in Unveil’s management environment. Support for third party systems – Conversation Manager 2.0 extends support of telecommunications systems and software from key partners Rockwell FirstPoint Contact, Nuance Communications and Avaya. To meet key customer requirements, the product integrates with Avaya’s MultiVantage ACD and Avaya Active Telephony CTI and integrates to Rockwell FirstPoint Enterprise. This integration enables Conversation Manager to perform Conversation Assists in a timely fashion, appropriately route calls and deliver rich customer information to agent desktops. Version 2.0 also includes all the recent functionality improvements in Nuance Voice Platform (NVP) version 3.0, providing a feature-rich and standards-based voice self service platform. Statistics and Report Server – Conversation Manager 2.0 provides an enhanced set of management reports that provide historical information on virtual agent staffing, performance, learning, and inquiry topics. Three report types are supported in version 2.0: • The virtual agent report provides an operational view on the performance Unveil’s virtual agents. This includes how many Virtual Agents were active and for how long, as well as how many assists and takeovers occurred. The virtual agent report simulates common reports on live agents a call center manager would obtain from a workforce management system. • The call summary report shows the ‘call view’ of system operations. This includes the average times calls reached a given stage; overall automation rates; call lengths; number of assists and number of takeovers that occurred. • The Conversation topics report provides a detailed analysis to the call center management of what customers are calling about. The Conversation topics report is the primary way of examining the conversational flow of the center. This includes the flow of calls into and out of a given topic, and statistics on time, turns, and assists within that topic. In addition to these key advances, Unveil Conversation Manager 2.0 features improvements in system scalability, enabling support for up to 200% more simultaneous virtual agents than version 1.2, advanced call control features to support conversation assists, routing and recording and increased performance of the product’s core Meaning Recognition accuracy. These improvements enable higher overall automation while significantly reducing caller perceived latency for call handling. Unveil’s Conversation Manager 2.0 is currently available to enterprise and outsourced contact centers in the North American market. www.unveil.com sales@unveil.com Jim Williams, Dir. of Marketing 781-250-1105 jwilliams@unveil.com

Vocomo Previews new voice-over-IP IVRProduct: Vocomo Corporation (Cupertinoa, CA) previewed its new open standards Voice-over-IP (VoIP) IVR platform: VocomoVoice® Response for VoIP. This Network Attached IVR platform, allows businesses to efficiently deploy their customer service applications in a single IP-centric network. VocomoVoice Response for VoIP supports the Session Initiation Protocol (SIP), H.323, VoiceXML 2.0, and speech technologies from IBM. It is scheduled for release on October 1st, 2004. www.vocomosoft.com or call toll free 1.800.780.8626. Contact: Vocomo Corporation Toll free: 1-800-780-8626 info @vocomosoft.com Danny Lange, CEO of Vocomo

Telephone - IVR Partnerships/Financials

Acapela answers the phone with Natural Touch: Natural Touch is a company specialised in developing and selling products and services based on telecommunication, internet and speech technologies, and it has chosen Acapela Telecom to answer its clients' calls with a smooth natural voice. With Dialoca Assistant, Dialoca Directory and Dialoca Software Suite, Natural Touch offers to take charge of and improve customer relations by identifying the caller's needs and giving each call personal attention. Waiting time and endless transfers are eliminated to provide efficient and valorizing communication. Acapela's speech synthesis can answer the phone calls, identify the requested correspondant and transfer the call, mentioning, if necessary, the absence of the correspondant when the call is redirected to the operator, or the answerphone. The navigation becomes natural, the dialog intuitive and efficient and speech synthesis becomes the natural answer to personalized dialog needs. Call processing solutions from Natural Touch offer a low cost, quality solution, fully adaptable to the client's needs, and have already convinced clients such as Darty, les Taxis Bleus, Paris Expos, and the French Ministry of the Interior. Natural Touch is also working on mobility applications : setting up solutions to follow people and vehicles on the go. www.naturaltouch.fr www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471 Natural Touch, Sevim Essiz 06 14 37 79 62 sevim.essiz@naturaltouch.fr Thierry Brizzi, president of Natural Touch

Audentify and European Technology Consultants Announce Alliance: Audentify (Cambridge, UK) a leading provider of next-generation contact center technology and a division of Autonomy Corporation plc , and European Technology Consultants Limited (ETC) a leading consultancy and systems integrator specializing in customer service and sales solutions announced a partnership that combines Audentify's leading technology with ETC's industry expertise to deliver highly effective customer contact management to customers throughout Europe. ETC provides solutions that increase the efficiency and effectiveness of an organisations customer service and sales capability. By forming a strategic partnership with Audentify, ETC strengthens its existing offering to customers. Audentify will form a key part of ETC's portfolio by forming a conceptual understanding of customer communications including telephone calls, email, Instant Messaging and web chat, understanding customers' needs and ensuring that they are dealt with in the most efficient way, thereby driving customer satisfaction. This helps organisations to compete more successfully, increase market share and improve gross margin. Audentify's speech recognition technology with sophisticated analytic capabilities and 100% recording facility accelerates agent productivity, efficiency, and, in turn, customer satisfaction. Unlike more traditional legacy systems reliant upon spotting key words, Audentify drives customer satisfaction by understanding the entire conversation, routing calls to experts and making proactive suggestions to the customer's enquiry. Audentify addresses the challenges faced by the modern contact center head on, such as increased call volumes, the need to decrease operating costs, improving agent productivity and raising standards of customer service through a unique range of functions which maximize efficiency and assist agents in their jobs in order to deliver real results to customers. Audentify, a division of Autonomy Corporation plc , is a leading provider of next-generation contact center technology. Built on Autonomy's unique Intelligent Data Operating Layer (IDOL), Autonomy Corporation plc is a global leader in infrastructure software for the enterprise. Autonomy's technology powers applications dependent upon unstructured information including call center, customer relationship management, knowledge management, enterprise portals, enterprise resource planning, online publishing and security applications. Autonomy's customer base includes more than 1,000 global companies including BAE Systems, Ford, Ericsson, Royal Sun Alliance, Sun Microsystems and public sector agencies including the US Department of Defense, NASA and the US Department of Energy. Strategic reseller and OEM partners include leading companies such as ATG, BEA, Business Objects, Citrix, Computer Associates, EDS, IBM Global Services, Novell, Novient, Veritas, Vignette, Supportsoft and Sybase. The company has offices worldwide. The Autonomy Group of companies includes: Aungate, a leading supplier of electronic communications management technology for regulatory compliance in the enterprise; Audentify, a leading supplier of next-generation contact center technology; and Virage, a leading provider of rich media communication and content management software European Technology Consultants Limited (ETC) is a leading consultancy and systems integrator providing customer service and sales solutions. ETC specializes in improving the efficiency and effectiveness of the customer interaction, addressing both multi-channel and individual contact channels - with a strong focus on web and contact centre. ETC's solutions include customer contact strategy and channel performance optimization across technology, business process and people. This helps organizations increase market share and gross margin, and gain competitive advantage. ETC provides management consultancy, systems integration and managed services to clients across the financial services, telecommunications and retail sectors. Founded in 1995, ETC has an enviable blue-chip customer base that includes Orange, Sony, Telewest and BT. ETC has strategic relationships with IBM, ATG, BEA and Genesys. Contact: Beth Ladd, Autonomy Corporation plc, +1-415/243-9955, bethl@us.autonomy.com Charlotte Herbert, +44(0)1223-448-000, charlotteh@autonomy.com Gary Pruden, ETC, +44(0)207-562-5100, gary.pruden@etc.co.uk

Avaya and IBM Collaborate: Avaya and IBM will jointly work on speech-enabled self-service solutions. The solutions - which are the result of an expanded alliance between the two companies - bring together IBM's business middleware and speech technology and Avaya's contact center experience. The solutions will combine Avaya's IP-based contact center software with IBM's WebSphere infrastructure software and speech technology, enabling enterprises to incorporate speech easily into a range of business processes: areas like account management, customer purchasing and inventory management, that are traditionally siloed away from the call center. This integration would then allow a business' customers and employees to use speech to interact with automated services that link together contact centers, IP networks and back-end infrastructures - resulting in improved operational efficiencies, customer satisfaction, and employee productivity. Built on open standards-based technologies, the joint offerings will enable integration of critical data such as updated customer information, past purchases and real-time inventory levels, ensuring automated services have all the information required to serve callers more effectively and with greater speed. Also, the companies assert that these offerings will let businesses extend enterprise-class speech self-service to any remote location across IP networks - including local and branch offices. For instance, when a banking customer makes an after-hours call to their local branch, callers will not be limited to voicemail, or be asked to call another phone number. The first phase of the alliance brings together Avaya's Interactive Response with IBM WebSphere Voice Server 5.1. New features in WebSphere Voice Server include WebSphere Application Server for common administration, installation, deployment and logging throughout the enterprise, as well as support for Red Hat Enterprise Linux WS/ES/AS for Intel 3.0 and SuSE SLES 8.0. Avaya will also support IBM's Reusable Dialog Components initiative, a move, with partners, to open-source a Java-based framework of pre-built application components for building speech applications. www.avaya.com www.ibm.com

Voxpilot and Kirusa to showcase Kirusa Voice SMS Multimodal Application deployed on Voxpilot’s carrier-grade Open Media Platform: Kirusa and Voxpilot demonstrated the patent pending Kirusa Voice SMS (KV.SMS) multimodal application, deployed on Voxpilot’s fully VoiceXML 2.0 compliant Open Media Platform, at this year’s SpeechTEK Conference in New York. The KV.SMS application, a combination of the two most popular mobile communication technologies available today, has been successfully tested in pilot projects by major European carriers. The application provides captivating features such as RingFree voice messaging, voice reply to text messaging, and text reply to voice messaging, greatly enhancing the reach of SMS messaging to a wider population of mobile users. Kirusa Voice SMS is useful in a wide variety of consumer and business situations, and does not require any network upgrades. Carriers are thus able to offer this dynamic service to drive increased data and voice revenue by leveraging their existing wireless network infrastructure. Voxpilot, a Eureka Soft company, delivers VoiceXML solutions for carriers, enterprises and service providers worldwide. Combining the proven performance of a carrier-class IN platform with cutting-edge VoiceXML technology, Voxpilot's open, standards-based solutions provide support for a broad range of operational and voice services that increase profitability and improve customer satisfaction. Voxpilot's VoiceXML Outsourcing Solution, which enables carriers and service providers to aggregate third-party applications on their networks, includes its widely-used Voxbuilder Online Development Environment, a graphical Web-based tool for the creation and management of VoiceXML applications. Voxpilot's customers include France Telecom, Wind, Cegetel, Monaco Telecom, Swisscom, Ericsson and Cofiroute. www.voxpilot.com. Kirusa, Inc. is a multimodal solutions company that has developed industry leading multimodal technology (with several patents pending) for the delivery of applications using multimodal interfaces. This next generation of wireless technology allows users to customize the presentation of information on wireless devices by either alternating, or simultaneously receiving information through visual and voice interfaces. Kirusa develops the unique Kirusa Voice SMS multimodal application, and supports third party developers of multimodal applications. Kirusa's customers include the worlds' leading telecom carriers, such as France Telecom, Bouygues Telecom, and Mobilkom Austria. +1 (732) 623-2118 or visit www.kirusa.com. Contacts: Dr. Michaela Heigl Voxpilot Tel: +33 6 62 67 01 69 Inderpal Singh Mumick Kirusa, Inc. Tel: +1 732 623-2118 ext. 217

SandCherry Resells Metaphor Solutions' Packaged Speech IVR Applications: Metaphor Solutions, and SandCherry, Inc. announced a partnership allowing SandCherry to resell Metaphor's suite of packaged speech applications. The alliance provides the products and services needed to develop, deploy and maintain speech solutions that enable wherever, whenever self-service to customers, employees and partners. Integrating Metaphor's pre-packaged speech applications with SandCherry's advanced voice services platform offers businesses cost-effective solutions that can be quickly implemented. The two companies will work together on market and customer development activities, targeting voice solutions in a range of industries including healthcare, financial services, retail, public safety, construction and government. Metaphor's comprehensive line of packaged applications for healthcare, finance, retail and enterprise applications will be re-sold as part of SandCherry's Voice4 industry-targeted solutions. AppPackage bundles offer comprehensive, affordable solutions that combine Metaphor's pre-built applications with the SandCherry SoftServer™ voice services platform and speech component software. Professional services are also included for installing and integrating all components in the customer's environment. The flexibility of SandCherry's Voice4 solutions allows companies to start with a single application and add capacity and new AppPackage solutions or custom applications as desired. SandCherry, Inc. is a communications software company that offers pre-packaged and custom speech application solutions, advanced software platforms, voice application development tools, and media components. SandCherry powers cost-effective touch-tone, speech-enabled and multimodal (data+voice) service delivery for enterprises and service providers. By extending existing Web applications with a voice user interface, companies can leverage their investment in web-based services and infrastructure to reduce the cost of delivering services to phone users while providing integrated multi-channel self service. 720-562-4500, toll-free 866-383-4500 www.sandcherry.com. Metaphor Solutions, Inc. offers packaged speech IVR (Interactive Voice Response) solutions to call centers and enterprises with the most human-like caller experience at the lowest cost of deployment. The company has the most number of packaged speech applications in the industry. The company delivers this through its patent-pending Metaphor Conversation Manager (MCM) that allows the fastest and most flexible speech IVR application development. This in turn allows small and medium enterprises to enjoy the large labor savings and superior caller experience of speech IVR with significant reductions in the cost, risk, complexity, and time to market. The company's packaged speech IVR applications run on standard voice gateways based on VoiceXML and SALT. Metaphor markets its applications directly as well as through a network of partners and resellers. Deployments can be hosted or installed on-premise. Callers can dial into Metaphor's inventory of speech applications and experience them live using the published numbers on the company website at www.metaphorsol.com. Contacts: SandCherry, Inc Robin Shively Phone: 720-562-4529 Email: rshively@sandcherry.com Michael Kuperstein, Metaphor Solutions, 781-239-0137, michael@metaphorsol.com

Speech Technology Associates Employees File Countersuit Against HTI Voice Solutions: Last month HTI Voice Solutions sued five former employees that had formed Speech Technology Solutions (STA) for allegedly conspiring against HTI Voice Solutions while working on opening a competing business. In the countersuit, the STA employees essentially denied all of the Counts in the HTI Voice Solutions suit that included: Breach of Contract; Unfair and Deceptive Practices; Breach of Covenant of Good Faith and Fair Dealing; Breach of the Duty of Loyalty; Defamation; Intentional Interference with Contractual Relations; Intentional Interference with Advantageous Business Relations; and Misappropriation of Trade Secrets. In their countersuit, the STA employees claim that HTI forced them to sign a "Confidentiality, Non-Compete and Non-Solicitation Agreement" and is not enforceable. It also claims that the HTI CEO (Ken Brater) "committed fraud and other egregious acts against HTI’e employees, shareholders and Board of Directors. Another counterclaim is that HTI has not sustained any damages as the result of the actions of the STA employees. www.speechtechassociates.com 978-692-5289 www.htivs.com

voice robots chooses teliSpeech, telisma’s speech recognition software for its new-generation voice portal platform: voice robots (Heidelberg, Germany) and Telisma (Paris, France) announced the conclusion of a worldwide retailer agreement. True to its quality-oriented philosophy, voice robots chose to integrate teliSpeech – telisma’s latest networked speech recognition technology - in its IVR platform robot 5 to increase its capabilities and give customers a wider choice of high-quality speech recognition. The new application suite will include auto-attendant, voice mail, calendar and name dialing modules. voice robots GmbH is a leading technology supplier in the voice solution market. The core voice technology platform, robot 5, is based on open standards and offers a unique GUI-based development suite for fast and easy speech application. development. voice robots and its partners deliver interactive automated voice solutions for telcos, ASPs and enterprise clients that address current business needs by reducing costs and increasing efficiency in business communication. Major companies such as Legion, a subsidiary of the Lagardère Group, Orange and Home Shopping Europe rely on the robust robot 5 platform to run their automated voice business. www.voicerobots.com Contact: voice robots, Erik Oosthuizen erik.oosthuizen@voicerobots.de Tel: +49 (6221) 505 401 Graham Lewis, voice robots CEO Telisma, Pascale Garreau pgarreau@telisma.com Tel: +33 (0) 1 42 45 45 91

Telephone Messaging – Applications/Products/Partnerships/Financials

AVST adds Nortel Switch Integration: AVST plans to offer a new digital integration between its flagship unified communications solution, CallXpress, and the Nortel Networks Business Communication Manager (BCM), an IP-enabled communication system with converged voice and data functionality for small and medium sized businesses. AVST also recently announced its CallXpress solution will be integrated with the Mitel 3300 Integrated Communications Platform. CallXpress effectively makes the cell phone an extension of the PBX, which enables customers who use the Nortel BCM or Mitel 3300 ICP in the office to use their cell phone as their office phone while traveling. www.avstgroup.com ontact Denny Michael, VP of Marketing at 949-699-2300 or.

AVST Closes $10 Million in Series B Funding: Applied Voice & Speech Technologies, Inc. (AVST) (Foothill Ranch, Calif.), a provider of second generation communications solutions for the enterprise market, obtained $10 million in new funding. Ignition Partners led the round with participation from Dolphin Equity Partners. Additionally, existing AVST investors have an option and have expressed a commitment to participate in the B round with an additional $2 million in investments. AVST plans to use the proceeds from this funding to accelerate product development and expand its sales and marketing efforts to support the increased demand for its innovative enterprise communications solutions. AVST has uniquely combined the strengths of its unified communications platform, CallXpress, with its innovative speech applications platform, Seneca, to create a second generation enterprise communications solution. With many companies seeking to replace their first generation technology over the next 24 to 36 months, CallXpress delivers a best-of-breed solution with its broad integration capabilities, scalability, modularity and extensive feature set including advanced voice messaging, unified messaging, and speech-enabled auto attendant and personal assistant capabilities. AVST has over 34,000 systems installed worldwide and an impressive list of customers, including many Fortune 500 companies, professional sports teams, healthcare facilities and government agencies. AVST also has a strong long-term OEM relationship with Ericsson which distributes CallXpress worldwide in 15 languages under the "OneBox" brand. With strong distribution partners including Verizon's Enterprise Solutions Group, Altura Communication Solutions and Norstan Communications as well as 150 additional independent resellers worldwide, AVST is poised for rapid growth. Cascadia Capital, a national investment bank specializing in the financing of companies in a number of market sectors including information technology and communications, assisted AVST in securing the financing. Ignition Partners, headquartered in Bellevue, Wash., is a premier venture capital firm focused on early stage computing, network infrastructure and telecommunications startups in the Western United States Ignition brings together an unparalleled combination of domain focus, technical expertise and operational experience. Ignition's partners are proven business leaders with more than 160 years of combined experience, who have built some of the world's most successful businesses of the last two decades, including Microsoft Windows, McCaw Cellular Communications, AT&T Wireless and Microsoft's enterprise business. www.ignitionpartners.com Dolphin Equity Partners, L.P., (www.dolphinequity.com) specializes in investments in emerging media, communications and related information technology companies operating in the U.S. and abroad. Located in New York, Dolphin currently manages over $340 million in capital. Cascadia Capital (www.cascadiacapital.com) offers corporate finance, mergers & acquisitions, strategic advisory services, and capital market services to companies across North America. Cascadia Capital serves emerging growth and middle market companies, both public and private. The firm focuses on the following sectors: Information Technology, Communications, Security & Defense, Healthcare, and Industrial & Consumer.Applied Voice & Speech Technologies, Inc. (AVST) is a developer of enterprise communications solutions for businesses of all sizes. AVST's solutions enable direct access to an enterprise's communications infrastructure including voicemail, email and fax messaging from any voice or data device via a touch-tone, GUI or speech interface. Established in September 2003, AVST was formed from the combined businesses of Sound Advantage (established in 1997) and the CallXpress division (established in 1982) of Captaris, Inc. Headquartered in Orange County, CA, AVST maintains facilities in Bellevue, WA and the United Kingdom and has remote sales offices across the United States. With over 34,000 systems installed worldwide, AVST's communications solutions are sold and supported internationally by an extensive network of resellers and OEM partners. Contact Denny Michael, VP of Marketing at 949-699-2300 or www.avstgroup.com.

Healthpoint Uses New Avaya Unified Communication Center to Increase Efficiency of Mobile Workers: One of the companies benefiting from Avaya's speech-enabled communications is Healthpoint Ltd., a growing global pharmaceutical company that is using the new version of Avaya Unified Communication Center to improve productivity of headquarters staff and field sales representatives. Healthpoint uses the application to let users have a single source where they can access e-mail, voicemail, calendars and other functions via speech. This greatly enhances mobility for workers in transit, who need their information rapidly and conveniently. In addition to tapping the software's abilities to allow Healthpoint's workers to use a phone to check or schedule dates in an online calendar, Shane says the company's sales force benefits from the "Reach Me" capabilities of Unified Communication Center. This gives Healthpoint's sales team greater access to customers, enabling them to remain reachable through just one number that will seamlessly track down employees, wherever they may be. Since its inception in 1992, HEALTHPOINT has established a U.S. presence in the research, development, and marketing of branded pharmaceuticals, over the counter drugs, and medical devices for tissue management, dermatology and surgical. An aggressive research and development effort has brought HEALTHPOINT's unique technologies to the market, setting the pace for its highly trained field sales organization of direct sales representatives. HEALTHPOINT, with over 290 employees, is a DFB Pharmaceuticals, Inc. affiliate company. HEALTHPOINT is based in Fort Worth, Texas www.healthpoint.com. Also headquartered in Fort Worth, DFB provides technology-driven products, outsourcing services, and licensing opportunities to the healthcare industry worldwide through its affiliate companies, contract partners, and branded marketing organizations. www.avaya.com contact: Jonathan Varman, Avaya 908-953-6432 jvarman@avaya.com

iVoice Provides Update on Company Streamlining and Five Common Stock Distributions: iVoice, Inc. (OTC Bulletin Board: IVOC - News) (Matawan, NJ) provided an update on the streamlining of the company created by the spin-off of five companies to shareholders. iVoice has determined that the best way to create shareholder value, separate and apart from the operating performance of iVoice, is to implement new business opportunities by distributing shares of spin-offs to the company's shareholders. The five common stock distributions are part of a broader strategy relating to the transition of iVoice into a company focused on the development and licensing of proprietary technologies. The company's recently developed and patent-pending Talking Pill Bottle is a good example of the types of assets that iVoice is retaining. iVoice will also continue its previously announced acquisition strategy. To fund these initiatives, iVoice has over $6.8 million in cash to date.

iVoice CEO, Jerry Mahoney, said, "We have taken major steps to restructure the company in ways that we believe will be most favorable to shareholders. Once all the transactions are completed, we will still own our core patent-pending technology, including the Talking Pill Bottle and the Voice Item Locator. The primary business activity of iVoice will be licensing this proprietary technology. We plan to file additional patents in the field of talking products in the near future." Mahoney noted that the company has recently hired GlynnTech to assist in facilitating various licensing opportunities. Mahoney continued, "The other major element in our strategy is finding and completing acquisitions that offer excellent growth potential, are in viable and stable market segments and have management teams committed to success. With over $6.8 million in cash on our balance sheet, we expect to have additional funds available after the closing of the sale of the speech engine and related intellectual property to NGM TEC." In February 2004, iVoice spun off Trey Resources, which closed on its first acquisition in June 2004 by acquiring an operating company with sales of more than $2 million. Trey subsequently announced a letter of intent to acquire an additional company. The successful Trey Resources spin-off demonstrated the feasibility of the mechanism and established further opportunities for both entities. iVoice shareholders received one share of Trey Resources for every 1,793 shares held in iVoice as of Feb. 9, 2004. On Sept. 7, 2004, iVoice announced the anticipated distribution to the company's shareholders of all common stock of a newly formed subsidiary, iVoice Technology, Inc. This move was designed to unlock the value in iVoice's interactive voice recognition (IVR) software technology by transferring this technology to an independent public company, iVoice Technology, and thereafter distributing the common stock of this subsidiary to the company's shareholders. iVoice Technology will initially focus on building a business around the IVR technology. Subject to final review by the SEC, holders of iVoice common stock will receive shares of iVoice Technology common stock on the record date, which has been tentatively set as November 1, 2004. The number of shares to be distributed to iVoice shareholders has not yet been determined. It is intended that the iVoice Technology common stock shares will trade on the NASD OTC Bulletin Board. On Sept. 9, 2004, iVoice announced a letter of intent to sell its speech engine and related intellectual property to NGM TEC, Inc., a provider of identity verification solutions based on its proprietary voice biometric technology. NGM TEC's product line addresses the rapidly growing problem of identity theft. As part of the transaction, iVoice will become a significant shareholder in NGM TEC. The transaction, which is subject to the completion of favorable due diligence, the signing of a definitive agreement and customary closing conditions, is expected to close in October 2004. The total purchase price of $6.25 million is payable in a combination of cash, debt and stock. iVoice intends to distribute the shares it receives from NGM TEC to individual iVoice shareholders via a special common stock distribution. This will give iVoice shareholders a meaningful stake in NGM TEC. The record date and the number of shares to be distributed will not be determined until the close of the transaction. NGM TEC, Inc. plans for its common stock shares to trade on the NASD OTC Bulletin Board. The foregoing description of this transaction is qualified in its entirety by reference to the Current Report on Form 8-K dated September 3, 2004, which was recently filed with the Securities and Exchange Commission. On Sept. 13, 2004 iVoice announced that it intends to distribute to the company's shareholders all common stock of a newly formed subsidiary, iVoice Technology 2, Inc., as a way to unlock the value in its Unified Messaging software technology. Unified Messaging links telephone systems to computer networks and allows users wherever they may be to access voice mail and retrieve e-mail in a single operation through existing local area networks. Subject to final review by the SEC, holders of iVoice common stock will receive shares of iVoice Technology 2 common stock on the record date, which has been tentatively set as November 1, 2004. The payment date will be as soon as practicable after the SEC declares the iVoice Technology SB-2 registration statement effective. iVoice Technology 2 plans to list its shares on the NASD OTC Bulletin Board. The number of shares to be issued to iVoice shareholders has not yet been announced. On Sept. 15, 2004 iVoice announced its fifth common stock distribution to shareholders of its entire holdings of 7.5 million Class A common stock shares of Corporate Strategies, Inc. to the company's shareholders. The record date for this distribution of Corporate Strategies common stock has been set for October 20, 2004 for shareholders of record of iVoice on that date. iVoice has invested $750,000 in Corporate Strategies in the form of 7.5 million shares of Class A common stock and a $500,000 5% secured convertible debenture. Corporate Strategies, located in Houston, Texas, provides merchant banking and diversified financial services involving accounts receivable factoring, mortgage lending, oil and gas investments and assorted other financial transactions. Corporate Strategies recorded sales in excess of $1.7 million in 2003 and pre-tax income in excess of $130,000. The foregoing description of this transaction is qualified in its entirety by reference to the Current Report on Form 8-K dated June 25, 2004, which was recently filed with the Securities and Exchange Commission. iVoice, Inc. designs, manufactures and markets innovative speech-enabled applications and computer telephony communications systems. Customers include ADP, Michael's Stores, BJ's Wholesale Club, American Red Cross and JVC. Through a single integrated solution capable of processing thousands of interactions per hour, the company's products provide cohesive, easy voice access to messaging systems, including telephone calls, emails, faxes and voice mail. Interactive Voice Response products developed by iVoice allow PC databases to be accessed via voice or from a standard touch-tone telephone.. A profile for investors on iVoice may be found at the website www.hawkassociates.com/ivoice/profile.htm. An online investor kit containing iVoice press releases, SEC filings, current price Level II quotes, interactive Java stock charts and other useful information for investors can be found at www.hawkassociates.com and www.hawkmicrocaps.com. Investors may contact Frank Hawkins or Julie Marshall, Hawk Associates, at (305) 852-2383, email: info@hawkassociates.com. www.ivoice.com Contact: Delores Serafin-Sloboda, investors@ivoice.com Jerry Mahoney, CEO, 732-441-7700, jerrym@ivoice.com

NGM TEC Terminates LOI with iVoice: iVoice, Inc. (OTC Bulletin Board: IVOC - News) announced that NGM TEC, Inc. terminated the Letter of Intent with iVoice due to an unanticipated delay in NGM TEC completing its initial round of financing.

Preferred Voice Releases Mobile Personalization Services Suite For Advanced User-Centric Communications: Preferred Voice, Inc. (OTCBB: PFVI), a provider of personalization services announced a new enhanced version of its Mobile Personalization Services Suite "My Phone Services" for wireless carriers that provides a new breed of wireless applications enabling user-centric communications. The suite includes tightly integrated products including the Advanced Network Address Book, Network Voice Activated Dialing, Push2Connect, SmartLine and Rockin' Ringback. Preferred Voice is one of the first companies to publicly offer a managed ringback service to wireless carriers in the U.S. Wireless carriers are increasingly called upon to offer personalized services that recognize the unique needs of their users. This user-centric vision provides the framework to deliver sustainable competitive advantage and is defining the next-generation of personalized services. Preferred Voice is redefining the personalization services landscape by taking a unique approach in its user-centric vision by tightly integrating its personalization services into a suite that enables users to simplify their growing number of products and enables the delivery of personalized services independent of devices or networks. As a result, users are able to define their own communications and entertainment experience unique to their needs while offering wireless carriers a meaningful and sustainable competitive advantage to improve customer retention and satisfaction and set high barriers for switching. The "My Phone Services" Mobile Personalization Services Suite is part of Preferred Voice's managed services for wireless carriers and includes the following integrated services: • Advanced Network Address Book • Voice Activated Dialing • SmartLine • Push2Connect • Rockin' Ringback The Advanced Network Address Book uses a personal database to manage information, services and preferences across a number of different applications. PVI taps into a powerful paradigm for user-generated content that includes its Advanced Network Address Book as the foundation for a family of services that create new revenue and retention opportunities for carriers. Studies have shown that users will pay the highest monthly subscription for address book contents, more than email/texts, calendar, or ringtones/games/graphics. Preferred Voice's Advanced Network Address Book offers a bundled solution that provides significant value for wireless carriers. The value proposition of offering a network solution is that it provides a secured back-up for the address book should the device be lost or broken. Voice Activated Dialing (VAD) is tightly integrated with the network address book as it stores the numbers to be dialed by voice. This addressable opportunity will grow as subscribers add more contacts to their handset and a growing number of users make it their primary phone. SmartLine provides advanced call control features enabling the user to decide what calls to allow all the time, depending on the day of the week or time of the day or not at all (call blocking). The integrated Push2Connect grouping feature enables users to personalize their contacts according to different criteria and enables an innovative one-touch conferencing and messaging to specific groups. PVI's new Rockin' Ringback is a network-based service that enables users to choose an audio file that callers will listen to while the phone is ringing. PVI's ringback service provides wireless carriers with a new way to increase average revenue per user (ARPU) while providing advanced customization and personalization services for subscribers. PVI's ringback service is compatible with any phone and across any network - wireless or landline. PVI's Rockin' Ringback service allows subscribers to personalize their dialing experience and entertain their callers with favorite music tracks, sound effects and self-created messages. For example, a subscriber could assign a ringback to her boyfriend so when he calls he hears Jessica Simpson's "Take My Breath Away" until she answers the phone or assign Will Smith's "Parents Just Don't Understand" to play for her parents when they call. One can also further select the tune to be heard on a particular day of the week and/or time of the day. Preferred Voice markets its products and services to wireless, ILEC and CLEC companies across the United States. Services include Safety Talk voice activated dialing, Emma TR the telephone receptionist, and Push2Connect conferencing on demand. Preferred Voice also provides carrier connectivity, call management, provisioning and service delivery solutions for carriers, voice portal providers, wireless application providers and other enhanced features providers. www.preferredvoice.com Contact: Joni Hanson, for Preferred Voice, Inc. (206) 972-6823 or joni@drivepointcomm.com Latif Nathani, CEO of Preferred Voice, (214) 265-9580

Pronexus Introduces VeoConnect - The Economical Speech-Enabled Auto Attendant for Microsoft Speech Server 2004 Intelligent Speech-Enabled Call Routing Delivers Caller Satisfaction and Optimizes Efficiency for Organizations of All Sizes Pronexus Inc.(Ottawa) announced the release of VeoConnect, its speech-enabled auto attendant for Microsoft Speech Server 2004. VeoConnect rapidly connects callers with their intended destination, enabling them to reach any department or employee by simply using their voice. This removes the often-frustrating dial-by-name experience and eliminates the need to remember extension numbers. This in turn leads to improved customer satisfaction and significantly reduced call-handling costs as calls are routed without human intervention. While speech-enabled auto-attendants are increasingly used by large organizations, VeoConnect is designed for the broader audience of small and medium enterprises. It delivers sophisticated features for caller satisfaction and ease of administration in a turnkey, economical application that makes speech accessible to many more organizations. -Dial by voice: Callers say employee names, departments or extensions to get connected easily. -Turnkey voice user interface with touchtone fallback and live operator options -Flexible speech recognition parameters. Options include recognition of first and last names, first or last name only, nicknames, department names, extension numbers spoken as individual digits or as whole numbers. -Advanced Call Routing: Regular-hours and after-hours call routing, holiday notification greetings, and company-wide announcements. -Online dictionary with correct phonetic information of many popular names. Dictionary is continuously updated and optimized by Pronexus' linguistic staff. -Built-in support for multiple recognition results - when the system has found more than one possible match to a caller's request, it present options to the caller so they can choose the intended destination. -Ease of administration: Secure web-based administration console for managing system settings and employee information -Detailed reporting capabilities - average call duration, hours of day when peak call volumes occurred, number of calls per hour, and much more. VeoConnect, part of the VeoSuite range of turnkey voice-enabled applications for Microsoft Speech Server, is now available from Pronexus. www.pronexus.com Contact: Sonia Langenberg (Tercler), Marketing Communications Pronexus Inc. Tel: (613) 271 8989 ext.541 sonia.langenberg@pronexus.com

IntelliSPEECH 4.2 Advances Reliability and Usability of Speech-Enabled Technology: System Development Company of New Hampshire, Inc. (SDC)(Manchester, NH) announced version 4.2 of IntelliSPEECH, SDC's speech-enabled call and information handling solution. Designed to further enhance user flexibility, customer self-service, and overall customer satisfaction, IntelliSPEECH 4.2 advances the reliability and usability of speech technology. An innovative speech solution, IntelliSPEECH 4.2 provides a comprehensive telephone directory database accessible via the spoken word. Inbound callers may connect their call, initiate a page, or access information via speech. IntelliSPEECH 4.2 eliminates fifty to seventy percent of routine calls, allowing operators to handle more important tasks and provide the highest level of customer support to callers requiring a personal touch. Updated features to IntelliSPEECH 4.2 include: *Emergency Out Calling - Using IntelliDESK Emergency Procedures, IntelliSPEECH 4.2 automates the process of emergency paging and auto-dialing. *Unique User Identification - Ensures system security by authenticating users before allowing system access or changes. Prohibits users from accessing and modifying information outside their defined user class. Designated users are able to override user class restrictions in order to reach an otherwise protected end point. *Email Reader - Enables callers to retrieve and reply to emails via speech. This application may be used with any email server that supports POP3/SMTP. *Custom Greetings - Enables different internal and external greetings. Abbreviated greetings may be programmed for use by internal speech users, while more detailed greetings may be offered to outside callers. In addition, external greetings can be changed on the fly and scheduled as needed from the telephone. Additional key enhancements and benefits include: *The ability to leverage existing web-based and/or custom applications *Enhanced, up-to-date grammars capable of handling the most complex speech recognition tasks *Improved call reporting to gauge and track effectiveness *Enhanced systems management and monitoring *Open standards supported architecture for diverse platform support *Standards-based deployments based on VoiceXML with a variety of IVR solutions IntelliSPEECH 4.2 is driven by Newfound Communications Advanced Grammar Management Kit, providing enhanced, up-to-date grammars to handle the most complex speech recognition tasks. www.newfoundcomm.net. IntelliSPEECH 4.2 also utilizes Vocalocity's VocalOS Voice Gateway, a VoiceXML 1.6 standard. The Gateway provides the coordination of multiple resources including telephony, voice recognition, text-to-speech, audio, media caching, VoiceXML and systems management. In addition, it seamlessly integrates disparate resources, allowing customers to use existing equipment, tie into existing legacy systems, and drive down costs by integrating components that best fit their needs. www.vocalocity.net. SDC is known for its superior solutions, technical support, and customer training and service. Through its IntelliSPEECH and IntelliDESK family of products, SDC provides speech recognition, custom CTI solutions, and PC-based intelligent consoles with on-line directory services for leading customers in the healthcare, public sector, education and corporate industries. SDC is a ScanSoft, Inc. (NASDAQ: SSFT - News) Business Partner. SDC is also a member of the Avaya (NYSE: AV - News) DevConnect Program the Cisco Systems (NASDAQ: CSCO - News) AVVID Partner Program, the NEC Fusion Program, the Nortel Networks (NYSE: NT - News) Developers Program and the Siemens (NYSE:SI - News) HiPath Ready Program. www.sdc-nh.com. Contact: System Development Company of NH, Inc., Robin Davis, 603-629-4242 rdavis@sdc-nh.com or Newfound Communications, Inc. James Loeber, 603-470-4797 jloeber@newfoundcomm.net or Vocalocity, Inc. Elyse Phillips, 650-969-3311 ephillips@vocalocity.net

UbiCall Communications Adds Languages to Voxplorer Virtual Receptionist: UbiCall Communications (Mons, Belgium), a speech products company, specialized in name dialing applications using Nuance ASR, has now released its Voxplorer Virtual Receptionist in German, German adapted for Switzerland, German and Brazilian Portuguese. It was due to the increasing demands from Switzerland, and Brazil, through the Ubicall partners and distributors in the countries, that UbiCall Communications decided to put some more priority on these versions. The next language in line will be Spanish, and it is planned for end of October. These languages are being added to following languages that UbiCall has been delivering off the shelf: US English, UK English, French, and Dutch. Voxplorer Receptionist is a cost-effective automated attendant with specific functionalities improving the smoothness of call transfer from caller to the requested party. Based on state-of-the-art Speech Recognition Technology and the highest standards of dialog design and user friendliness, Voxplorer Receptionist is available for small enterprises, medium-sized organizations and large corporations. UbiCall offers the VoxPlorer Receptionist on either an ASP or a CPE basis. UbiCall partners include: Cisco, Alcatel, Avaya (Switching partners) and with Genysis, Envox, Voxpilot as platform vendors. Customers include: Belgian Ministry of Finance, Belgacom, Dexia Bank, Ethias NRB, Several Hospitals out of the I.R.I.S group, Delhaize LME Business Center. University of Fribourg, Opel/Saab Motor Willems, and SBMI. http://www.ubicall.com/ Contact: Willy Van de Voorde, Director of Sales, Ubicall Communications sa/nv, Mobile: (+ 32) (0) 475/ 24 44 90, Direct Nr: (+32) (0) 65/32 15 62 Office Tel: (+32) (0) 65/32 15 46 email: willy.vandevoorde@ubicall.com

Operator Services – Applications/Products/Partnerships/Financials

ScanSoft and varetis Power Fully Automated Directory Assistance Service for Telix AG: ScanSoft, Inc. (Nasdaq: SSFT) and varetis announced that they have worked together to deliver easy-to-use automated directory assistance to the German telecommunications company Telix AG. The service is available via the short code DA number 11810. Based on the ADIS directory assistance platform from varetis and using the ScanSoft Directory Assistance application, the system enables callers to quickly and easily retrieve all of their directory assistance requests - from phone numbers to addresses - via a fully-automated voice user interface. It provides fast, accurate real-time access to more than 40 million subscriber listings throughout Germany. In addition, varetis ADIS gives consumers the option of making inquiries from their wireless devices via SMS, enabling Telix to reach a new, untapped demographic of younger DA users. The combination of technology from varetis and ScanSoft results in a state-of-the-art solution to meet customers' demand for more information and faster and greater cost efficiency in directory assistance services, resulting in a short, smooth and informative conversation. Just as customers would speak with a live operator, the dialog asks for the required information in an easy-to-use and fluent manner, without any hold times. For a successful inquiry, the system only needs the location and either company or last name. For residential systems, the first name and street name of the requested person may be given and if more than one listing is found, the system presents the alternatives and allows the caller to select the desired listing. The Telix project is a great example of bringing the right partners together for the job, and it signals a milestone for DA voice automation. With 40 million entries, the German national database is the largest DA database in Europe and provides a unique, multi-level approach to input and dialog," said Christian Mooser, head of voice automation at varetis. The varetis ADIS solution is the result of over 10 years of research and development in the specialized field of voice automation for directory assistance. Backed by varetis' patented, powerful search algorithms, ADIS allows callers to interact in a completely natural way - quickly and without confirmations. The voice-automated service is available in Germany at 0.49 Euro per minute. Over the last two decades, varetis AG (Frankfurt: WKN 691 190) has equipped more than 240 telephone service centers and nearly 18,000 agent positions worldwide with its products and service. Around 160 people work at the headquarters in Munich and at regional sales offices in the UK and Singapore. Today, its software powers the DA services of nearly all of the world's major carriers. www.varetis.com Telix AG has been offering nationwide directory services in Germany since 1999 and employs 35 professionals in Berlin and Munich. Under the short code 11810, it recently expanded its portfolio to include fully automated inquiries via either voice or SMS - a revolutionary first for Germany. www.telix.de Contact: ScanSoft International Vanessa Richter, +32-9-239 80 31 vanessa.richter@scansoft.com or ScanSoft North America Jonna Schuyler, 978-977-2038 jonna.schuyler@scansoft.com or varetis AG Anja Meyer, +49 (0)89-455 66 186 Fax: +49 (0)89-455 66 313 anja.meyer@varetis.de Kristian Pazdersky, CEO at Telix AG

varetis and telisma join forces to promote a new generation of automated directory assistant at competitive costs for telecom operators. varetis (Munich, Germany) and telisma announce the availability of telisma’s technology – teliSpeech - on varetis Directory Assistance system, ADIS. ADIS (Automated Directory Inquiry System) is the fully voice-automated directory assistance solution from varetis, built upon a number of proven varetis products. ADIS provides fast and accurate voice access to large telephone directories, processing even ambiguous recognition results within minimal response times. At the same time, ADIS presents a consistent and natural user interface to callers for an excellent user experience. By automating voice inquiries, ADIS frees up call center agents to process the more complex inquiries and higher-revenue calls that are characteristic of enhanced directory services. In addition, ADIS comes with revenue-generating features such as call completion and SMS messaging, and its operator Fallback capabilities ensure that each call is processed correctly. Varetis chose teliSpeech for its ability to answer the very stringent requirements of its ADIS national directory assistants. According to Christian Mooser, Head Of Voice Automated Solutions, varetis "When it comes to mass-market automated directories, speed is the key to success. People simply won’t tolerate having to wait – even half a second. The challenge is to find the right telephone number, name or address within milliseconds from among up to hundreds of millions of listings. TeliSpeech’s introduction and support of Extra Large Vocabularies, as well as its ability to compile extra large grammars in real time and online is of particular interest to us in these applications." Varetis also selected teliSpeech for the quality and size of its European language library and its ability to handle multilingualism and regional accents. High recognition accuracy is made possible by teliSpeech’s intrinsic support of phonetic variants and proper-noun lexicons. For national directories, the technology can recognize all kinds of regional pronunciations and even typical mispronunciations by non-native speakers. In countries where several languages are spoken – as well as countries with a high proportion of non-native speakers - real multilingualism allows for an efficient way of handling foreign utterances. Commercial considerations also entered into the creation of this partnership. ( www.varetis.com) (www.telisma.com) A software editor, telisma is European leader in Automatic Speech Recognition technologies. Telisma carrier grade products empower mass market speech activated services with a unique obsession for accuracy and dialog fluency. Created in 2000, telisma is a private company headquartered in France, which also operates from local offices in London and Munich. varetis AG Anja Meyer, +49 (0)89-455 66 186 Fax: +49 (0)89-455 66 313 anja.meyer@varetis.de Christian Mooser, Head Of Voice Automated Solutions, varetis

 

 

Telephone Tools – Applications/Products/Partnerships/Financials

Brooktrout Supports Microsoft Speech Server with its Telephony Interface Manager: Brooktrout Technology, Inc., announced its Telephony Interface Manager (TIM) for Microsoft Speech Server. Qualified software vendors and enterprise customers interested in developing their speech applications on the Microsoft Speech Server platform can join Brooktrout's Early Access Program to start developing on the platform. Brooktrout is collaborating with Intervoice, a world leader in converged voice and data solutions to integrate the Intervoice Omvia Voice Framework on Brooktrout's TR1000 speech platform. In addition to its work with Microsoft, Brooktrout collaborated with Intervoice to ensure that packaged and custom applications written to the Intervoice Omvia Voice Framework will work seamlessly on the TR1000 for Microsoft Speech Server. Intervoice will also be offering CTI and VoiceXML add-on modules that will be compatible with Brooktrout's TR1000 platform to provide customers with a growth path to more advanced call center features. www.brooktrout.com/. Brooktrout Technology is a subsidiary of Brooktrout, Inc. . Contact: Laura Often Lida Joly +1 781-453-4858 +32 2 658-0190 loften@brooktrout.com lida@brooktrout.com

Eicon Networks Announces New System Software and Toolkit Releases for Its Diva Server Family of Telephony Adapters: Eicon Networks Corporation ("Eicon") announced the availability of new releases of Diva Server System Software Suite and Diva Server Software Development Kit (SDK). Diva Server Software Suite version 7.5 provides a common system software and driver package for the whole family of Diva Server adapters -- including Analog, ISDN BRI, ISDN PRI, and E1/T1 telephony adapters. With Diva Server SDK version 3.4 software developers and system integrators gain a powerful and easy-to-use toolkit to port their communications solutions to the cutting-edge Diva Server telephony adapters as well as the new Diva Server SoftIP software product to facilitate the migration towards IP Telephony. As a major addition to the Diva Server family of products, Diva Server Analog adapters are being enabled by the new version 7.5 of Diva Server Software Suite. Thus any communications application once developed for a digital Diva Server adapter can immediately benefit from the new Diva Server Analog. Specific analog call control is abstracted in such a way that standard CAPI or TAPI programming interfaces can be used seamlessly. With the support of color fax Diva Server Software Suite version 7.5 offers a major feature enhancement for all CAPI based fax applications. For voice recording and conferencing applications Diva Server Software Suite version 7.5 offers new features to suppress DTMF tones, to automatically adjust the recording volume and to set the sampling rate to arbitrary values. Diva Server Software Suite version 7.5 is available for the most recent Microsoft Windows and Linux based operating systems. Drivers for Microsoft Windows 2000, Windows Server 2003 and Windows XP, including 64-Bit editions, are signed and certified by Microsoft's Windows Hardware Quality Lab (WHQL), further easing hardware installation. Drivers for the most recent Linux distributions of Red Hat, SuSe and Debian are readily available for download from the Eicon Networks web page. For other Linux distributions and kernels Diva Server Software Suite version 7.5 is available as Linux source-level RPM, facilitating porting Diva Server applications to any Linux-based operating system. Offering a high-level abstraction of the most frequently used call control and media streaming functions as C/C++ programming libraries as well as ActiveX controls for Visual Basic and VB.Net environments, Diva Server SDK version 3.4 makes applications development easy and straight-forward. As a major enhancement Diva Server SDK version 3.4 includes support for Diva Server SoftIP, a new software product that extends Diva Server call control and media streaming into IP Telephony, using industry standard Voice over IP (VoIP) and Fax over IP (FoIP) signaling protocols. Using Diva Server SDK version 3.4 in combination with Diva Server SoftIP, any communications application once developed for Diva Server can immediately be used in IP-Telephony environments, talking to IP Phones, PSTN/IP Gateways and IP PBX systems. Diva Server SDK version 3.4 offers a comprehensive set of new features further streamlining application development in the area of voice, speech and conferencing. Feature highlights include additional means to control audio streams for efficient integration with 3rd party speech recognition engines as well as additional audio formats including G.726, GSM and ADPCM (6 kHz and 8 kHz sampling) based voice coding. For managing conferencing applications Diva Server SDK version 3.4 also supports more sophisticated means to handle member rights, defining listen only, speak only and supervisor members. Furthermore, Diva Server SDK version 3.4 offers a complete suite of functions for developing passive monitoring applications. Both, recording of control and audio channels over an ISDN BRI and PRI interface are available using any standard Diva Server adapter in high-impedance mode. Diva Server Software Suite version 7.5 and Diva Server SDK version 3.4 will be commercially available in October 2004. Eicon Networks is a world-leading provider of reliable communication products for networked business applications, secure remote access and virtual private networks. The company's portfolio includes a full range of products for applications in branch office and PC secure connectivity, voice, speech and conferencing, voice over IP, POS and financial clearing networks, unified messaging and fax serving. Eicon distributes, markets and supports its products in more than 80 countries through a network of channel partners, telecommunication companies (telcos), Internet Service Providers (ISPs), and original equipment manufacturers (OEMs). www.eicon.com. Contacts: Kipton Heuertz, V.P. Product Marketing, Americas Eicon Networks Tel: 972-473-4530 kip.heuertz@eicon.com Leigh Ann Clavenna Marketing Manager, Americas Eicon Networks Tel: 972-473-4533 leigh.clavenna@eicon.com

Envox Worldwide Announces Support for ScanSoft SpeechPAKs: Envox Worldwide announced that it has expanded its strategic relationship with ScanSoft to include support for ScanSoft SpeechPAKs, pre-packaged components that can be used to create speech solutions. As a result, Envox 6, the company’s open, standards-based communications development platform, is now tightly integrated with ScanSoft’s VoiceXML-based SpeechPAKs. Additionally, the Envox Professional Services organization has adopted SpeechPAKs as a critical tool in their development of speech-enabled voice solutions. The new SpeechPAK capabilities will allow the Envox Professional Services organization and Envox 6 customers to more quickly develop and deploy speech-enabled solutions and to base those solutions on best practices in user interface (UI) design. Envox’s integration support for the ScanSoft SpeechPAKs is the latest example of how the two companies are working together to reduce the time and cost of developing voice solutions. Envox Worldwide and ScanSoft have also collaborated to provide Automated Directory Assistance solutions for the telco market and comprehensive speech-enabled, VoiceXML development systems for voice solution developers. Envox 6 is an open, standards-based platform with integrated software components that significantly reduce the time, cost and complexity of developing and deploying voice solutions. Envox 6 provides unmatched flexibility because it supports the widest array of communications and IT technologies and standards in the industry, including speech technologies, VoiceXML, host media processing, VoIP, and Web services. It empowers developers to create a wide range of voice solutions and to reduce development time by 50% or more. The Envox 6 value proposition is built around the ability to leverage existing investments in hardware, software and solution development, while providing a smooth migration path to the emerging standards and enabling technologies that will be required in the future. The Envox Professional Services organization provides comprehensive consulting and development services, based on Envox’s unique blend of IT and communications expertise, to ensure that our customers achieve their voice solution objectives. Services provided include solution development, user interface design, application integration, tuning, and deployment services. The company delivers its solutions through a global network of channel partners, including VADs, VARs, Systems Integrators, Service and Solution Providers, ISVs, OEMs, ASPs, consultants and other resellers. www.envox.com. www.ScanSoft.com. Envox Media Information, contact Gretchen Clarke, +1 508 871 7603, gclarke@envox.com

IBM to Contribute Speech Software to Apache Software and Eclipse Foundations IBM announced it is contributing software to the open source community in a move to spur the availability of speech-enabled applications by making it easier and more attractive for developers to build and add speech recognition capability in a standardized way. The initiative, supported by more than 20 key industry players from speech vendors to platform providers, is aimed at ending the battles over competing, proprietary specifications. IBM is contributing Reusable Dialog Components (RDCs) to Apache Software Foundation and proposing a project at the Eclipse Foundation to donate markup editors for speech standards established by the W3C. Pre-built speech software components, or "building blocks" that handle basic functions such as date, time, currency, locations (major cities, states, zip codes), RDCs are often-used functions in speech-enabled infrastructure applications. These allow a caller to, for example, book a flight using an auto-agent over the phone. Multiple reusable dialog components can be aggregated to provide higher levels of user functionality. Developed by IBM Research, RDCs are Java Server Page (JSP) tags that enable dynamic development of voice applications and multimodal user interfaces. JSPs that incorporate RDC tags automatically generate W3C VoiceXML 2.0 at runtime -- providing a standard basis for speech applications. By providing familiar and standards-based programming models, J2EE developers can add voice interaction to Web applications. And by making the RDC framework available to the community, speech components built using it will work together, regardless of the vendor that created them. Both the framework and a set of example tags are to be contributed to the Apache Software Foundation. Separately, IBM's contribution of speech markup editors to Eclipse is aimed at making it easier for developers to write standards-based speech applications as well as create and utilize RDCs within those applications. In proposal stage, this contribution not only gives speech developers a standard way of writing VoiceXML applications, it can also give web developers tools to more easily add speech access to their web applications. This comprises the initial formation of a project at Eclipse for open source tools for voice application development, which will be further evolved by several companies in the VoiceXML community. Supporters of this initiative include: Apptera, AT&T, Audium, Avaya, Cisco, Fluency, Genesys, Kirusa, Loquendo, Motorola, Nortel, Nuance, Openstream, ScanSoft, Siebel, Syntellect, Telisma, TuVox, V-Enable, Viecore, Vocomo, VoiceGenie, Voice Partners, and VoxGeneration. Currently, much of the code in the speech ecosystem is proprietary and specific to each vendor. This initiative is aimed at giving speech developers the benefits of open, standards-based programming models and tools that mainstream developers have had. This can also allow companies to speech-enable their existing applications more quickly and efficiently since developers will be able to build speech applications from standards-based components from various speech providers in the same application. "Since its initial $40 million contribution to launch Eclipse in November of 2001, IBM has continued to contribute to making Eclipse an open platform for application development and integration," said Mike Milinkovich, Executive Director of the Eclipse Foundation. "With this project proposal, IBM is taking another step toward propelling innovation and giving Java developers the tools to work speech technology into their applications." The initiative follows closely on the heels of IBM's contribution of its Cloudscape database to the Apache open source community as well as additional developer resources for the community to build on cloudscape. "This is the latest step in IBM's contribution to open source and to speech technology," said Gary Cohen, General Manager, IBM Pervasive Computing. "By giving more standards-based speech resources to the development community, IBM hopes to accelerate development and drive innovation in all areas of the speech ecosystem -- from speech vendors, to ISVs, to platform providers." IBM is the world's leading e-business company offering a wide range of services, solutions and technologies that help businesses take full advantage of emerging innovation. IBM's pervasive computing and wireless strategy is to extend e-business applications to the new class of client devices. This involves building and implementing wireless applications, creating groundbreaking initiatives to set open industry standards; and deploying hundreds of wireless consultants. IBM has the ability to develop, integrate and deliver solutions to meet the requirements of Network Operators, Broadband Services Providers and municipal/regional government business and infrastructure requirements. The Apache Software Foundation provides organizational, legal, and financial support for a broad range of open source software projects. The Foundation provides an established framework for intellectual property and financial contributions that simultaneously limits contributors' potential legal exposure. Through a collaborative and meritocratic development process, Apache projects deliver enterprise-grade, freely available software products that attract large communities of users. The pragmatic Apache License makes it easy for all users, commercial and individual, to deploy Apache products. For more information on the Apache Software Foundation, please see www.apache.org/. Eclipse is an award-winning, open source platform for the construction of powerful software development tools and rich desktop applications. Leveraging the Eclipse plug-in framework to integrate technology on the desktop saves technology providers time and money by enabling them to focus their efforts on delivering differentiation and value for their offerings. Eclipse is a multi-language, multi-platform, multi-vendor supported environment that it is built by an open source community of developers and it is provided royalty-free by the Eclipse Foundation. Eclipse is written in the Java language, includes extensive plug-in construction toolkits and examples, and can be extended and deployed on a range of desktop operating systems including Windows, Linux, QNX and Macintosh OS X. Full details on Eclipse and the Eclipse Foundation are available at www.eclipse.org. IBM open-sources speech-recognition development tools: IBM announced today that it will release as open-source code some of its software for speech-enabling applications. The move is intended to spur development in the field and outflank rivals by making its free technology the industry standard. IBM is donating to the Apache Software Foundation its Reusable Dialog Components (RDC) technology, a set of building-block components for adding basic, voice-enabled functions to applications like data, time, currency and location interactions. IBM's RDCs are Java Server Pages tags; IBM is donating to Apache its framework and a set of example tags. The company is also donating a set of speech markup editors to the Eclipse Foundation, the guardian of the open-source Eclipse integrated development environment IBM created. www.ibm.org

Loquendo Releases VoxNauta Lite 6.0 Software Speech Platform: Loquendo, the global speech technology company, has released the latest version of Loquendo's Speech Platforms: VoxNauta Lite 6.0. VoxNauta Lite 6.0 is the result of Loquendo's long experience in providing high-quality speech technologies and minimizing the customer’s service deployment effort. It is a complete multi-lingual open solution for building customized voice service and improving communication through voice access to information – anywhere, anytime. VoxNauta Lite 6.0 supports the latest releases of Loquendo ASR 6.x, Loquendo TTS 6.x, Loquendo Speaker Verification, and W3C VoiceXML 2.0 Recommendation. Its renewed packaging allows seamless upgrade to forthcoming enhancements of speech technologies, new voices and languages. The new VoiceXML 2.0 fully compliant interpreter provides additional tags for Speaker Verification along with a number of enhancements ensuring fast reaction to VoiceXML service interpretation. VoxNauta Lite 6.0 offers different service profiles spanning from simple IVR with DTMF and prerecorded audio, up to full dialog ASR, TTS, and Speaker Verification. New service profiles can be defined according to customer needs. The customer can exploit VoiceXML programming potentialities even with entry-level profiles based on simple touch-tone services. In a second time, ASR technology may be added, with total reuse of the developed voice services. VoxNauta Lite 6.0 provides great flexibility in terms of network interoperability: it supports analog and digital switched network connections, as well as VoIP based on our SIP/RTP protocol implementation. Both NMS and Intel boards support is ensured. The platform is complemented by Loquendo Studio 6.0, a suite of tools assisting the service developer in any stage of the service deployment: from grammar editing and tuning, up to VoiceXML service debugging and testing. With over 30 years R&D expertise, Loquendo is at the forefront of the global speech technologies market. Its high-quality, high-performance technologies (Loquendo TTS - text to speech, Loquendo Embedded TTS, Loquendo ASR - automatic speech recognition, Loquendo Speech Server, Loquendo Speaker Verification) and platforms guarantee systems integrators the best solutions in 15 languages and 34 voices - with more coming soon. www.loquendo.com

Loquendo and NMS Communications Certify Interoperability Under NMS Universal Speech Access - Industry's Only MRCP-Enabled Speech Interface: NMS Communications (NASDAQ: NMSS) announced that Loquendo's Automatic Speech Recognition (ASR) software based on the popular Media Resource Control Protocol (MRCP) 1.0 protocol has been certified by NMS as compliant with the NMS Universal Speech Access interface. The NMS interface is the only enabling technology product currently available for MRCP 1.0, which is widely endorsed as a de facto standard by today's major speech engine vendors. NMS' Universal Speech Access provides developers with a common application programming interface (API) offering standards-based access to today's most in-demand speech resources. Loquendo's NMS certification, as a result, means that development issues caused by traditional proprietary interfaces are eliminated, resulting in simpler integration, significantly lower lifecycle costs, and quicker time-to-market for application developers. NMS and Loquendo have cooperated on delivering conventional 'audio provider' interfaces in the past. With NMS' Universal Speech Access, Loquendo has now added MRCP support for its prominent Loquendo ASR engine, and will shortly provide MRCP support for the Loquendo TTS engine, both of which will continue to drive Loquendo's leadership in the fast-growing speech applications space. Part of NMS' Open Access(tm) family of products (http://www.nmscommunications.com/OpenAccess), Universal Speech Access helps alleviate one of the major development challenges inherent in writing speech applications - the difficulty using multiple types of speech resources efficiently. With Universal Speech Access, developers are able to write applications once, under a common API, and then access any speech resource supporting the MRCP standard. Additional resources, even from other vendors, can be used with minimal adjustments and no specific dependence on operating systems, versions, or other components - saving valuable time and money. NMS demonstrated Universal Speech Access at the recent SpeechTek trade show in New York, where it was evident that application developers are continuing to add automatic speech recognition (ASR), text-to-speech (TTS), and other speech capabilities to their products at a rapid pace. www.loquendo.com www.nmscommunications.com Contact: Ornella Ambrois Marketing Manager, +39 011 757 6850 Ornella.Ambrois@LOQUENDO.COM

MAP Telecom launches "VoiceXML Business Exchange" or VBXÒ in October 2004: MAP Telecom is pleased to announce the launch of a new service in the VoiceXML ecosystem: VoiceXML Business Exchange" or VBX. VBX is a meeting place for companies that are active in the voice and telecom world. Content and solution providers, solution demanders, ASP's, Telco’s and service providers will find each other on a forum and networking platform. The products clearing house will provide clients with up to date information about products and services that interest them, while their business will be known to potential customers and partners among MAP Telecom’s network of over 1600 expert developers. Media, Marketers and content-providers can use MAP Telecom offer of services on a daily bases in order to be dynamic in their market, monitor applications and generate revenue. This is a free service that MAP Telecom will supervise and which will bring more business opportunities to all its clients and partners. www.map-telecom.com/vbx/ MAP Telecom created the VoiceXML ecosystem in 2002 with the acquisition of one of the most performing and compliant VoiceXML service platforms in the market and operating the first online VoiceXML development environment (Voxpilot platform is one of the few having successfully passed the very demanding VoiceXML Forum certification tests). www.voxpilot.com In December 2002, MAP Telecom, with industrial partners Eurekasoft and Dexem, created Passport VoiceXML, the first independent VoiceXML competence centre, organizing professional training for developers entering the VoiceXML arena. www.passport-voicexml.com www.eurekasoft.fr and www.dexem.com In September 2003, MAP Telecom launched VUI Producer, helping application developers to produce their Voice User Interface without investment in voice talents or expensive speech synthesis licenses. www.vuiproducer.com MAP Telecom is an international voice service operator covering more than 26 countries. We provide a full set of infrastructures and services that achieve excellence in the following domains: - A VoiceXML platform, hosted in a telco-grade environment. - A fully deployed voice / data telecom network. - Number management and financial services such as micro payment. - A business exchange program with an ecosystem of more than 1600 developers www.map-telecom.com Contact: Marketing & Communication Manager, Nadine Daguenet, Email: n.daguenet@map-telecom.com Tel: +377 97 77 33 33

Metaphor Solutions Announces High-level Speech Application Development Tools: Metaphor Solutions (Wellesley, MA) announced the release of its Metaphor Conversation Manager (MCM). Metaphor Conversation Manager is an integrated development environment for developing automated speech applications that interact with callers, with back-end data sources and web services and with live service agents. MCM is as easy to use as writing a service script which allows unprecedented speed in development. It works with any VXML gateway and the Microsoft Speech Server. Metaphor Conversation Manager is an editor, linker, debugger and run-time interpreter that dynamically generates open-standard scripts in Voice XML and SALT from a high-level design-scripting language called Metaphor Script. It also includes Web-based end-user customization as well as administration, logging and reporting tools. MCM is 5-10 times faster than developing speech applications in straight Voice XML or SALT. It is 3-4 times faster than developing with graphical based Voice XML form objects with property sheets. This is because Metaphor Script, the high level scripting language of MCM, leverages many inherited resources used in the best practices for dialog design and has much more flexibility in controlling the conversation flow, as well as a simulation debugger. MCM includes a library of complete and fully tested package applications that allow the developers to quickly customize and integrate speech IVR solutions for most common applications. Also included are an array of reusable and fully configurable dialog building blocks that help users easily and quickly build both speech and touch-tone voice applications that are customized to their business operations. The MCM integrated development environment incorporates best practices in dialog and voice application design so that users get high-quality applications in a rapid timeframe, reducing development cycles and associated costs. It allows customers to integrate back-end systems to share legacy and Web information with their Speech IVR platform. The complete MCM package helps businesses easily transition from proprietary IVR and fully take advantage of the benefits of VoiceXML, SALT and Speech IVR. Open and standards-based, MCM-developed applications are VoiceXML and SALT compliant and eliminate the lock-in associated with proprietary IVR systems. www.metaphorsol.com. Contact: Michael Kuperstein, Metaphor Solutions, 781-239-0137, michael@metaphorsol.com

Newfound Communications Offers Limited Promotion on Nuance Voice Platform Starter Kit for $9,999 US: Newfound Communications announces a new promotional price of $9,999 for the Nuance Voice Platform (NVP) Starter Kit. The NVP Starter Kit is a turnkey and affordable VoiceXML 2.0 solution for developers and small/medium-sized enterprises (SME). Offered in conjunction with Nuance Communications (NASDAQ: NUAN) and Paraxip Technologies, the $9,999 promotion price is for a 4-port system. Available in 4-, 12- and 16-port configurations, the NVP Starter Kit is preconfigured with Nuance Voice Platform 3.0, the Nuance Application Environment, the software-based Paraxip Gateway, a rack mount speech server from Alliance Systems and analog telephony hardware from Intel. The NVP Starter Kit also includes the Nuance 8.5 speech recognition engine,; and Nuance Vocalizer 3.0 text-to-speech software. Applications such as auto-attendants, flight information, account balance look up and voice activated dialing have helped major enterprise and telecommunication carriers to reduce costs, increase customer satisfaction and create new revenue streams. Now those same benefits can reach the growing SME marketplace with the NVP Starter Kit. Newfound Communications, Inc. offers technology, tools, and professional services to fuel the speech recognition revolution. www.newfoundcomm.net or call 1-86-NEWFOUND.

Pronexus and Vail Systems Offer Developers "A SIP of Speech": Pronexus Inc. and Vail Systems, two Microsoft Speech Server partners and providers of telephony and speech solutions, announced the availability of 'A SIP of Speech', a unique software bundle aimed at developers exploring the growing opportunities for speech applications. The combination of Pronexus VBSALT and the Vail SIP TIM on the Microsoft Speech Server platform lets developers rapidly create and test speech applications from a single workstation or a workstation/server combination, all without requiring any telephony hardware investment. In combination with one or two softphones, this effectively creates an end-to-end environment for speech application development and testing. 'A SIP of speech' allows developers to leverage existing programming skills and seamlessly connect to SIP-based IP telephony networks while evaluating the potential for speech in their respective organizations. 'A SIP of Speech' is now available for download from the web sites of Pronexus and Vail Systems Inc. For additional information on pricing and both companies' range of tools, solutions and services for the Microsoft Speech Server, visit www.pronexus.com or www.vailsys.com/mss Founded in 1993, Pronexus Inc. is a leader in Computer Telephony and Speech. It is the developer of the award-winning VBVoice, and the new VBSALT, Rapid Application Development tools for business-critical telephony and speech solutions. Pronexus also provides VeoSuite, a range of turnkey speech applications for Microsoft Speech Server and a complement of professional services that voice-enable a variety of industries and applications. Comprehensive support and acclaimed training complete the firm's offerings. Delivering enhanced telephony solutions for over 10 years, Vail Systems is the reliable choice in speech. The Vail SIP Voice Platform is a proven, carrier-grade distributed computer network that processes a wide range of interactive voice services. These services include on-demand audio conferencing, outbound notification, and other speech-enabled IVR applications, and VoiceXML and SALT application hosting. Vail also offers a suite of packaged speech application solutions and the Vail SIP Telephony Interface Manager (TIM) for Microsoft Speech Server. (800) 360-8245 Contact: Sonia Langenberg, Marketing Communications, Pronexus Inc., Tel: (613) 271 8989 ext.541 sonia.langenberg@pronexus.com Joseph Carey, VP of Marketing, Vail Systems, Tel: 1-800-360-8245 jcarey@vailsys.com

Vicorp Announces Launch of xMP Narrator: Vicorp announces xMP Narrator, an innovative GUI-based design tool that enables Voice User Interface Designers to create speech applications with rich and flexible speech interfaces modelled around ‘natural conversations’ that take place between two people. xMP Narrator represents an important breakthrough in the evolution of speech services because it provides a new approach to Conversational Modelling that overcomes many of the technical limitations imposed by proprietary design tools. xMP Narrator will make speech applications more intuitive to use; and easier and less frustrating to design. Recent improvements in speech recognition and text-to-speech (TTS) quality have lead to the emergence of speech interfaces that callers find more natural to use. Unfortunately, many of the ‘natural language’ VUI design tools that exist today tie Developers into a complete tool suite that may not be suitable for their overall needs. xMP Narrator is designed to address this. It models the interaction between the ‘caller’ and the ‘system’ as a series of natural ‘conversations’ between two people - and then guides the VUI designer through the process of creating a series of conversational alternatives. Once captured, these conversational alternatives are generated into formal requirement documents, which in turn are used to specify services. Text written for the system side of the conversations then forms the basis of transcripts for recordings or Text to Speech renderings. A Revolution in Conversational Modelling ‘Recording transcripts’ can be produced for these prompts, and actual media files automatically incorporated back into the system. These are then used to create an open, standards-based content archive that can be independently deployed and managed by 3rd party tools, thus eliminating ‘lock in’ to any manufacturer or solution. Alternatively, the Vicorp model can be used to automatically generate a call flow, within xMP Director representing the conversation and allowing for all the captured content and meta-data to be provisioned for the service, including grammars, content, TTS and transcripts using this tool. xMP Narrator integrates with other tools in the Vicorp xMP suite as well as providing a cross-platform content archive. xMP is Vicorp counts many of the world’s largest Telecommunications Service Providers and Enterprises amongst its customers. www.vicorp.com . Scott Sibuns, VP Worldwide Sales for Vicorp.

Vocalocity Adds Text-to-Speech Capability: Vocalocity will be increasing the features offered through its telephony offerings with text-to-speech capabilities from NeoSpeech. The two companies announced a partnership at a conference in New York. NeoSpeech develops speech applications for handheld, desktop and network devices. Vocalocity says the NeoSpeech capability will add "crisp, realistic voices" for speech applications to its "voice browser". The company develops software and tools based on open standards. It is privately held with funding from Sutter Hill Ventures, Five Paces Ventures, Imlay Investments, Georgia Tech's ATDC Seed Capital Fund, and Atlanta Technology Angels. www.vocalocity.com Contact: NeoSpeech public relations, Jennifer Deitsch, 925-831-2092 jennifer.deitsch@neospeech.com or Vocalocity, Kim Martin, 813-263-4849 kmartin@vocalocity.net

VoiceObjects to open its voice application server architecture to the global voice industry: VoiceObjects(Cologne, Germany) announces VoiceObjects XDK.. VoiceObjects XML Development Kit will allow service creators to equip any voice or DTMF application with an interface to VoiceObjects’ industry leading voice application management system (VAMS) that combines high-performance voice application server capabilities and comprehensive management tools. With VoiceObjects XDK, organizations using any service creation environment will be able to run their voice or DTMF application on a powerful and highly scalable voice application server. VoiceObjects XDK will enable the industry to manage any existing or newly built application covering the whole lifecycle from deployment, monitoring, analyzing, and maintenance up to the re-launch. By demonstrating successfully implemented voice applications, the company’s experts highlight the advantages of server-based management systems with an integrated development environment (IDE). Live demonstrations, and customer references exemplify that VoiceObjects Server significantly reduces the total cost of ownership for newly developed and already existing voice applications due to centralized administration and lifecycle-management. Current customers comprise the infotainment voice portal of T-Mobile and telephone-banking applications as adopted by Postbank, Germany’s largest retail bank. The software simplifies and accelerates the realization of manifold telephone-services. The concept of VoiceObjects liberates Voice User Interface Designers from programming and gives them creative freedom for the design of user-friendly Voice User Interfaces with a high standard of intelligent dialogs. Functional elements for multilingualism and personalization add up to achieve better operability and user acceptance. With recently founded VoiceObjects, Inc, customers are now served out of the Boston area. VoiceObjects develops and markets VoiceObjects Factory, a leading Voice Application Management System (VAMS) for fast and easy provision of telephone-controlled services. VoiceObjects Factory accelerates voice application development and deployment. It enables higher service quality through built-in life cycle management, and allows for seamless integration of existing Web and back-end systems. With the investments from SAP Ventures (SAP) and T-Venture (Deutsche Telekom), and TeleSoft Partners, VoiceObjects has global players on its side. The company maintains a worldwide network of strategic partners such as Comverse, Genesys, IBM, T-Com, T-Systems, and VoiceGenie. Current customers comprise companies such as Postbank (Germany’s largest retail bank), T-Com, and T-Mobile among others. www.voiceobjects.com/ Contacts: VoiceObjects AG (Headquarters), Sabine Winterkamp, Phone.: +49 2204 845-142 swinterkamp@VoiceObjects.com VoiceObjects Inc. Todd Coutu, Phone: +1 781 685 4608 tcoutu@VoiceObjects.com http://www.voiceobjects.com/

Voxpilot first to pass VoiceXML Forum certification tests:, Voxpilot’s (Dublin) Open Media Platform, which supports leading industry standards such as SS7, SIP, MRCP, SSML and VoiceXML, has now been independently certified for its conformance to the W3C VoiceXML 2.0 Recommendation. This is the result of rigorous conformance testing carried out earlier this year by the VoiceXML Forum, which has since then only successfully tested and certified two other platforms in the industry. Voxpilot will continue to support the further development of W3C standards, and ensure that its Open Media Platform remains at the forefront of standards-based communication technology. Voxpilot, a Eureka Soft company, delivers VoiceXML solutions for carriers, enterprises and service providers worldwide. Combining the proven performance of a carrier-class IN platform with cutting-edge VoiceXML technology, Voxpilot's open, standards-based solutions provide support for a broad range of operational and voice services that increase profitability and improve customer satisfaction. Voxpilot's unique VoiceXML Outsourcing Solution, which enables carriers and service providers to aggregate third-party applications on their networks, includes its widely-used Voxbuilder Online Development Environment, a graphical Web-based tool for the creation and management of VoiceXML applications. Voxpilot's customers include France Telecom, Wind, Cegetel, Monaco Telecom, Swisscom, Ericsson and Cofiroute. Contact: Dr. Michaela Heigl, VP Marketing & Communications, tel.: +33 6 62 67 01 69. The VoiceXML Forum is an industry organization formed to create and promote the Voice Extensible Markup Language (VoiceXML). With the backing and contributions of its diverse membership, including key industry leaders, since its inception, the VoiceXML Forum has successfully driven market acceptance of VoiceXML through a wide array of speech-enabled applications. Currently, the Forum has more than 325 member companies, distributed across three membership categories (Sponsor, Promoter and Supporter). Membership in the Forum is open to any interested company. The core activities of the VoiceXML Forum are driven by its four active committees: Education, Marketing, Tools and Conformance. The Forum also publishes the VoiceXML Review, an electronic magazine dedicated to VoiceXML, featuring insightful articles written by industry experts. Its public Web site (www.voicexml.org) a rich, one-stop source for information about VoiceXML and related technologies. The VoiceXML Forum is organized as a program of the IEEE Industry Standards and Technology Organization (IEEE-ISTO). www.voicexml.org.

Embedded Telematics – Products/Applications/Partnerships/Financials

CarBot , integrated in-car aftermarket entertainment system, speaks Acapela! CarBot (Canoga Park, CA), a company devoted to the creation of the best in-car entertainment platform, has chosen Acapela high quality text to speech to voice enable its in-car computer. CarBot allows anyone to bring their digital music and MP3 collection into the car, and have up-to-date e-mail and news read while driving. CarBot is an in-car entertainment computer system. It can be installed into any car that has a radio. CarBot uses Acapela high quality text to speech to read back names of song folders, names of songs and artists, while navigating the MP3 folder hierarchy. The entire system can be used with a handheld remote control and voice prompts - no screen required. Acapela TTS reads back e-mail or any other text that is sent to the unit via WiFi or wireless connectivity. www.carbotpc.com www.naturaltouch.fr www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471 Carbot, Damien Stolarz , CEOdamien@carbotpc.com 818-919-9493

PrairieComm Partners with ART: PrairieComm, Inc., a developer and supplier of 3G-baseband processors and communications software, announced it has partnered with ART Advanced Recognition Technologies to enable integration of the smARTspeak product line into PrairieComm’s FronTier 3G reference phone platforms. ART software adds a complete suite of speech recognition capabilities including: speaker-independent and speaker-dependent name dialing, continuous digit dialing, and command and control, to PrairieComm’s FronTier 3G reference phone platforms. The agreement allows FronTier customers to incorporate either smARTspeak, smARTspeak NG, smARTspeak XG, or smARTspeak XGT into their handset design. In addition, PrairieComm and ART will work to integrate the smARTspeak product line into the appropriate user interfaces for the FronTier application framework. Betty Finkelstein Marketing Communications ART, Advanced Recognition Technologies Ltd. Kiryat Atidim, Building No.2 Tel: +972 3 7685018, Mobile: +972 64 700018 Mailto:betty@artcomp.com www.artcomp.com/

First Russian TTS engine for Symbian: Sakrament company announced the release of Sakrament TTS Engine Symbian edition. This solution is the first Russian speech synthesis software for Symbian, the best-in-class operating system for mobile devices. With the rapid growth of smartphone industry, voice solutions for data-enabled mobile phones quickly expand mobile communications capability. For the first time in the history of the mobile market there is a possibility to use synthesized Russian speech on Symbian-powered mobile devices in all kinds of services and applications, from reading aloud SMS messages to creating voice-enabled personal organizers and other software. One of the technologies developed by Sakrament allows the smartphone to make programmed calls using synthesized speech which helps the manufacturers and solution providers to meet a broader range of user requirements. Symbian OS is quickly becoming the industry standard for next-generation mobile devices. More than 6 mln. Symbian smartphones have been shipped worldwide, and by the end of 2004 the total number of data-enabled mobile phone models will exceed 20. Together with Windows CE and other mobile platforms, developing voice solutions for Symbian is one of the top priorities for Sakrament. Sakrament was founded in Minsk, Belarus, in 1994. Sakrament is leading developer of voice solutions for Russian language. The company’s main products are Sakrament TTS (Text-to Speech) Engine for PC, PDA, and Symbian platforms (the PC engine is 100% compatible with Microsoft SAPI 5.1.), Sakrament ASR (Automatic Speech Recognition) Engine, Sakrament Personal Voice Master, a unique program for creating your own synthesized voice, and a number of other solutions in the sphere of voice recognition and synthesis. Sakrament has established partnership with such recognized brands as HP, Microsoft, IBM, Alcatel and many others. Learn more about Sakrament at www.sakrament-speech.com Symbian was founded in Great Britain in 1998 by Ericsson, Nokia, Motorola and Psion with the aim to develop a single OS for next generation mobile phones. Symbian OS enables the developers to use a set of standard APIs and develop universal mobile applications. Today Symbian is supported on more than 15 mobile phone models (Nokia 3660/3620, 6600, 7610, 7650, 3650/3600, 9290, 9210, N-Gage, N-Gage QD, Siemens SX1, Sony Ericsson P800, P900, Sendo X, Motorola A920, A925). Learn more about Symbian OS at www.symbian.com Contact: Sakrament. Alexander Ostapenko, Sakrament company, tel.+375 172 840650 ales@sakrament.com

Infineon, Volkswagen roll-out 'basic' telematics platform: A new telematics platform rolled out by Volkswagen and Infineon Technologies AG seeks to provide a scalable platform for in-car communications and other telematics features at the same price as a standard telephony platform. The partners said at the recent International Automotive Electronics Congress that the new platform's low price could help speed acceptance of telematics by car makers by deploying the technology beyond high-end vehicles. The new platform, dubbed "Basis-Telematik-Einheit" (basic telematics unit), will be promoted to device manufacturers, automotive OEMs and tier-one suppliers. The adoption of a standard telematics platform, especially in high-volume, mid-range car models, could significantly reduce the cost of the car networks, the companies said. Since platform functionality varies by software configuration, different manufacturer could generate unique features under the proposed telematics scheme. Basic functions would include a Bluetooth connection between the driver's cell phone and the platform. Hence, the integrated device could also make use of speech recognition for dialing, transmit car error messages to the manufacturer or a repair shop and provide location-based services using a built-in GPS module. The platform is built around a 32-bit controller based on Infineon's Tricore familiy. It also includes a GSM mobile radio module, Infineon's SingleStone Bluetooth module and a GPS navigation chip set also developed by Infineon, based here. The controller runs under embedded Linux, but any available operating system could be ported to other systems used by OEMs, said Marcus Venmann, marketing manager for Infineon's infotainment product unit. Beyond the operating system, Volkswagen developed all the software. The modular architecture stresses reusability and open interfaces. Another major criteria was eliminating the use of proprietary software, a VW spokesman said. The new telematics platform will compete against the one jointly introduced by Microsoft and Italian car maker Fiat in July. In comparison to the Fiat-Microsoft platform, the VW-Infineon system is more autonomous since it does not rely on existing automotive electronics. The "Microsoft [and Fiat platform] uses more external resources," Venman said. Also, the use of open standards and Linux makes the Infineon-VW platform look more like an open system. Volkswagen will test the unit internally, beginning later this fall. The first systems won't hit the road for another one or two years, the partners said, since OEMs will have to complete additional software development and integration before the telematics system is deployed.

Embedded Appliances – Applications/Products/Partnerships/Financial

iVoice Announces Patent Application Filing on Wireless Methodology for Talking Consumer Products
iVoice, Inc. (OTC Bulletin Board: IVOC - News), announced it has filed a new patent application with the U.S. Patent and Trademark Office. The application is titled "Methodology for Communicating Product Use Instructions via Wireless Technology." The invention relates to users hearing spoken instructions from product packaging by pressing a button. iVoice has so far received two speech recognition patent awards and has additional applications for speech-enabled telephony applications pending, including the filing announced today. The invention includes methods for communicating product use instructions to a consumer by wireless communication from a central processor to a plurality of product containers for subsequent audio speech playback. A method of communicating product use instructions to a plurality of consumers comprises providing a plurality of product containers with a microprocessor attached to each product container and an activation button for initiating the talking instructions. iVoice announced in August that it has received notification of a patent allowance award for the company's "Hear I Am!" Voice Locator technology. "Hear I Am!" Voice Locator enables a shopper to pick up a phone anywhere in a store and ask for the location of an item. The locator instantly reports the aisle, department or other location of the item. For example, a customer can pick up a phone and say "Milk." The locator will respond with "Milk is located in aisle seven between the butter and the bread." iVoice has entered into a technology licensing agreement with GlynnTech Inc. to serve as its licensing agent for speaking product packaging technology. www.ivoice.com Contact: Delores Serafin-Sloboda, investors@ivoice.com Jerry Mahoney, CEO, 732-441-7700, jerrym@ivoice.com

New York Metropolitan Transportation Authority (MTA) selected Loquendo Embedded Technology for the new M-7 cars produced by Bombardier: Loquendo's embedded text-to-speech technology has been selected by MTA Long Island Rail Road and MTA Metro-North Railroad. The Loquendo software will be integrated into the Pocatec communication system onboard the new M-7 fleet supplied by Bombardier Transportation. This will allow text messages issued from the Railroads remote facilities to be broadcast through the audio and visual information system on board railcars. Loquendo is the only speech technology vendor that provides a complete product line for servers, desktop, multimedia, automotive, PDA and embedded devices, guaranteeing the same wide range of high-quality voices and languages, and the same core engine in all these environments. The Loquendo text-to-speech synthesis engine uses the "Unit Selection" concatenative technique on a wide range of sound samples. With its continuing new voices production Loquendo consolidates its leadership in the latest generation of voice synthesis technologies and guarantees efficiency features such as, portability, natural timbre, intonation, and pronunciation accuracy. With over 30 years R&D expertise, Loquendo is at the forefront of the global speech technologies market. Its high-quality, high-performance technologies (Loquendo TTS - text to speech, Loquendo Embedded TTS, Loquendo ASR - automatic speech recognition, Loquendo embedded ASR, Loquendo Speech Suite, Loquendo Speaker Verification) and platforms guarantee systems integrators the best solutions in 15 languages and 34 voices - with more coming soon. www.loquendo.com Established in 1974 and a consistent market leader, Pocatec designs, manufactures and supplies advanced information and entertainment systems for the Mass Transit Industry . Pocatec is ISO 9001 and CMM Level 2 certified. Pocatec is the premier manufacturer of integrated audio and visual systems. Pocatec boasts a solid reputation for both it's products and service. www.pocatec.com Contact: Ornella Ambrois Marketing Manager, +39 011 757 6850 Ornella.Ambrois@LOQUENDO.COM

Opera browser to recognise speech: Opera Software is to include IBM's embedded speech recognition technology ViaVoice in the next version of its Web browser. Web users will be able to navigate, request information and complete Web forms by speaking, which should offer added convenience as more users access the Web on smaller, mobile devices, Opera said. "Voice is the most natural and effective way we communicate. In the years to come it will greatly facilitate how we interact with technology," said Christen Krogh, Opera's vice president of engineering. "By making this technology available today for the Wider Web audience, the serious work of voice-enabling the Web can commence." According to Opera, Web developers can add voice input and output to traditional HTML-based pages by using the XHTML+Voice (X+V) markup language. In a swipe at Microsoft, Opera said the technology could even let users replace PowerPoint: using the Opera Show presentation tool, users could create a multi-page presentation which they could show one page at a time using the voice command "page down," without approaching the computer. Igor Jablokov, director of embedded speech at IBM and chairman of the Voice XML Forum, tthe technology can help Web users start interacting with content in a more natural way, combining speech with other input- output tools. Opera bills its browser as faster, smaller and more standards-compliant than its rivals. The Norwegian firm recently debuted on the Oslo stock exchange and sold around 25 per cent of its share base for $35m. www.opera.com +47 24 16 40 00 Jon S. von Tetzchner, CEO

Cingular Launches New Software to Meet the Communications Needs of its Blind and Low-Vision Customers: Florida Customers Now Have Access to a 'Talking Phone' BOCA RATON, Fla., Cingular Wireless today announced the launch of TALKS, a speech-enabled application designed to address the needs of blind and low-vision mobile device users. TALKS by Cingular Wireless gives the blind and low vision consumer equal access to many of today's advanced wireless services. Created in partnership with ScanSoft, Inc. (Nasdaq: SSFT), TALKS is compatible with the Nokia 6620, essentially turning it into a "talking mobile phone." TALKS converts menus, instructions and content displayed on the Nokia 6620 mobile phone screen into speech for output to the phone's internal speaker or through an optional wired or Bluetooth headset. According to the American Foundation for the Blind, there are approximately 10 million people who are blind or have low vision in the United States.* TALKS allows the blind and low-vision community to be more productive since audible cues allow users to access advanced communications services such as text messaging, email and phone directories. One of the unique features of TALKS is the talking incoming caller ID via natural sounding, computer- generated speech. Typically, blind and low-vision mobile users must answer the phone to learn who is calling. TALKS eliminates the guessing game by enabling a person to hear who is calling and be selective with the calls they take, as would any sighted customer. "The American Foundation for the Blind applauds Cingular for its innovative approach to addressing cell phone accessibility for people who are blind or visually impaired," commented Paul Schroeder, vice president, program and policy group, American Foundation for the Blind. "This is the first advanced cell phone released in the U.S. that allows users to access all of its many features through speech output. We commend Cingular for its leadership in ensuring that all their customers can take advantage of state- of-the-art communications technology." "TALKS levels the mobile playing field by giving our blind and low-vision customers the ability to access and interact with some of the most advanced wireless products and services available today," said Rich Guidotti, vice president and general manager, Cingular Wireless-South Florida. With TALKS and the Nokia 6620, users can**: - Hear caller ID information and check logs of incoming and outgoing calls - Write and hear emails, text messages and notes - Manage contact information and dial phone numbers from the directory - Use the appointment calendar - Edit and add contacts, settings and profiles according to their personal preferences - Use the calculator, unit converter, alarm clock and other tools - Receive audible information about the battery level, network and signal strength. TALKS uses ScanSoft's software to make the wireless handset talk by providing dynamic, detailed and personalized information through an automated voice familiar to the blind community. TALKS will tell users if email or text messages have arrived or how many calls they missed, and will speech-enable the most important key functions for the phone. TALKS allows blind and low- vision customers to fully integrate today's wireless technologies into their lives. The product "package" includes the handset and software, sold separately. The Nokia 6620 can be purchased for $299 after the $100 rebate from Nokia; this is based on a two-year commitment. The TALKS software retails for $199, however Cingular Wireless is offering an introductory rebate offer. For a limited time, customers with qualifying visual disabilities can receive a $199 service credit rebate (for a two-year contract) or a $100 service credit rebate (for a one-year contract). Currently, the Nokia 6620 is the only handset Cingular offers that is compatible with TALKS. There are three ordering options for TALKS by Cingular Wireless: - Order by calling the National Center for Customers with Disabilities (NCCD) at 1-866-241-6568. The NCCD is available from 7 a.m. to 9 p.m. ET, Monday through Friday and on Saturday from 9 a.m. to 6 p.m. ET. The NCCD can also provide installation assistance by phone www.cingular.com TALKS is listed as an accessory item for the Nokia 6620. - Order at a Cingular Wireless retail store. An associate can assist in the ordering process. TALKS by Cingular will be delivered to the customer. www.cingular.com/disabilityresources Cingular Wireless, a joint venture between SBC Communications (NYSE: SBC) and BellSouth (NYSE: BLS), serves more than 25 million voice and data customers across the United States. A leader in mobile voice and data communications, Cingular is the only U.S. wireless carrier to offer Rollover, the wireless plan that lets customers keep their unused monthly minutes. Cingular is the largest provider of GSM services in the United States, and has a 100 percent digital GSM/GPRS footprint across its service area. It also launched the world's first commercial deployment of wireless services using Enhanced Data rates for GSM Evolution (EDGE) technology. Cingular also provides corporate e-mail and other advanced data services through its GPRS, EDGE and Mobitex packet data networks. * Statistic from American Foundation for the Blind can be found at www.afb.org ** Some capabilities of TALKS software cannot be used without the addition of optional Cingular services such as text messaging and multimedia messaging. These services are available, at an additional charge. CO: Cingular Wireless; SBC Communications; BellSouth; National Center for Customers with Disabilities; NCCD; TALKS; ScanSoft, Inc.; American Foundation for the Blind; AFB; Nokia

ScanSoft Expands Portfolio of Text-to-Speech Software for Embedded Applications with RealSpeak Solo 4.0: ScanSoft, Inc. (Nasdaq:SSFT) announced the availability of RealSpeak Solo 4.0, the latest offering in its industry-leading portfolio of text-to-speech solutions optimized for embedded and desktop applications. RealSpeak Solo 4.0 is the first text-to-speech product released post-merger from the SpeechWorks Division of ScanSoft and draws on some of the most popular high-quality voices from the company's broad network speech offerings, providing a highly scalable, complete embedded TTS solution. SpeechWorks Division of ScanSoft's TTS has proven to be a highly accurate solution within a wide range of leading customers including Peiker, Auralog, Ford Motor Company, NCR and Volkswagen. With memory sizes of 8-16MB for embedded and 30MB and 60MB for desktop, RealSpeak Solo 4.0 delivers premium voice quality and natural-sounding synthesized speech and can be incorporated into a wide range of applications, including: -- An array of consumer devices such as handsets, talking dictionaries, gaming consoles and desktop solutions. -- Automotive applications including hands-free dialing, navigation systems and command and control. Optimized for embedded navigation applications, RealSpeak Solo provides accurate pronunciation of navigation related text commonly used in automotive solutions and navigation devices. -- Assistive devices and technologies including talking automated teller machines, language learning solutions, screen readers and other assistive technology devices. RealSpeak Solo 4.0 Features -- Based on RealSpeak, the most widely used TTS engine available, RealSpeak Solo shares the front end interface, allowing for intelligent and contextual handling of ambiguous text input such as abbreviations and numbers -- Highly scalable and customizable, the RealSpeak Solo SDK is available for Microsoft Windows (Win98, WinME, WinNT, W2K and XP) and Microsoft WinCE Version 3.0 and higher, and includes 11, 16 and 22kHz versions, enabling developers to choose a maximum quality for their hardware configuration -- RealSpeak Solo is SAPI 5.0 compliant, and comes with an easy-to-use proprietary ScanSoft API with support for Speechify Solo API for an easy transition for existing Speechify customers -- RealSpeak Solo 4.0 brings the best of the former Speechify and RealSpeak voice portfolios under one umbrella, and ensures investment protection by providing backwards compatibility and a seamless upgrade path for customers. The SpeechWorks Division offers a full suite of market-leading TTS solutions for embedded and network-based TTS applications that, depending on the technology, support open, industry standards such as SSML, MRCP and SAPI for empowering VoiceXML- and SALT-based applications among others. The company's network-based TTS solutions provide customers with the most comprehensive and flexible solutions available today. Its embedded TTS solutions are ideal in mobile, automotive, game and PC/multi-media environments that require memory footprints ranging from 100KB to 16MB, and 30 and 60MB versions. Customers can also take advantage of SpeechWorks' Virtuoso Custom Voice Program to create a custom voice for their embedded or network-based TTS application, which can also be specifically tuned for a dedicated application domain, such as navigation. www.ScanSoft.com Contact: ScanSoft Jonna Schuyler, 978-977-2038 jonna.schuyler@scansoft.com

Interactive Language Training on a Chip: Sensory Inc, is now demonstrating its new patent-applied-for English language training technology that is implemented on low cost chips. The new technology is being introduced in Hong Kong at Sensory’s "Designing a Voice User Interface" seminar. The Chinese government has mandated that all service employees will learn English by the 2008 summer Olympics in Beijing. This has caused a rapidly growing market to emerge for handheld language training devices. Typically, these language trainers will use chips from companies like Sensory or Scansoft to play samples of the English phrases and to allow a Chinese speaker to compare his/her recorded voice with that of the English speaker. Sensory’s new technology goes one step beyond this to perform an analysis of the Chinese speaker’s phrases so the speaker can get immediate feedback on the pronunciation. It’s much like an English instructor on a chip! A phrase is recorded into a handheld device and the technology analyzes the phrase’s phonemic content, which is compared with that in stored templates of the properly spoken phrase. Playback of the recorded phrase can be sped up or slowed down, and an LED or LCD display can provide feedback on the quality of the spoken phrase during playback in order that the user can visually assess the performance. Settings can be adjusted to switch between a "lenient" instructor and a "tough" instructor, and training phrases can be cycled through to find specific phrases or general areas to practice (e.g. shopping, business, socializing, etc.). Sensory has chosen to rollout and showcase its language training technology on the RSC-4x Family. This family of integrated circuits (IC’s) has the ability to support a wide variety of speech technologies on a general purpose 8-bit microcontroller core. The RSC-4x Family is supported by a full suite of tools including, assembler, C-Compiler, debugger, and voice coding tools (QT2SI vocabulary generator, Quick Synthesis speech compressor, etc.). Edgar Chau, Managing Director of product development house Cyber-Workshop says "We are familiar with the RSC-4x Family of IC’s, and we see much demand from our customers for this new language training technology. The Beta unit we received from Sensory worked quite well and we are sure to have many designs for 2005". Sensory plans to bring its language training technology to other embedded platforms, including the ARM line of microcontroller products popular in cell-phones. Sensory has created a demo platform that showcases a product design concept for a low cost language-trainer. Sensory is seeking strategic partners to roll out English language training handheld products around the world. The complete system is expected to retail for $19.95-$49.95 depending on features. Sensory, Inc., based in Santa Clara, CA, is the world leader in embedded speech technologies. Sensory’s speech technologies include language training, speech recognition, speaker verification, speech synthesis, and animated speech, and are used in consumer electronics, cell phones, PDA’s, Internet appliances, interactive toys, automobiles, and other applications where low cost and high quality is essential. Sensory offers a complete line of integrated circuit and embedded software solutions to customers such as Hasbro, JVC, Kenwood, Matsushita, Mattel, Mitsubishi, Sharper Image, Sony, Toshiba, Uniden, and many others. Sensory is a profitable privately held company founded in 1994. www.sensoryinc.com Contact: Sensory, Inc. Tom Tolbert, (408) 240-1577 marcom@sensoryinc.com

Wizzard Software Announces Canadian Patent Application Filing for talking Pill Bottle: Wizzard Software Corp. (OTCBB: WIZD) announced it has filed a patent application through its MedivoxRx Technologies, Inc. division with the Canadian Patent and Trademark Office. The application is titled "Pill, Capsule, or Medicine Container Audio Instruction Systems and Methods for Providing Audio Instruction s with Pill, Capsule or Medicine Container." 2003 was a record year in the growth of retail prescriptions filled by Canadians. There were over 361 million prescriptions filled in Canadian retail pharmacies in 2003 (excludes hospitals). Retail prescriptions grew 7.9% over 2002, the largest annual increase during the last decade. Last year's retail prescription market totaled $15.9 billion, which represents on average 11 prescriptions per Canadian at an average retail price of $44 per prescription. It follows that Canadian physicians were just as busy in 2003 as their health care colleagues in pharmacy, especially those in family practice. There were 309 million visits made to office-based physicians last year and a family physician/general practitioner was seen in 73% of these visits. Of the more than 361 million retail prescriptions filled in Canadian retail pharmacies in 2003, 80% of all prescriptions (new prescriptions and refills) were written by family physicians. Family physicians comprise about half of the approximately 57,000 doctors in Canada. Rex, the talking medication bottle, assists pharmacies in complying with stringent prescription medication regulations by having the dosage instructions and other critical information read "out-loud" to patients in addition to providing the instructions on the standard label. This audio assistance helps to ensure that a patient fully understands what the medication is, how often to take it and when to refill the prescription or discontinue its use. Canada is the world's seventh largest market in terms of pharmacy sales (purchases by retail pharmacies) at US$5.5 billion and it is the second fastest growing. In addition to assisting pharmacies with their compliance initiatives, Rex helps meet the needs of an aging population as people lose their sight, cognitive functionality and physical abilities. Rex allows a patient to simply push a button on the bottle and hear their name, prescription type, dosage instructions, drug warnings and other information read in a clear, friendly voice. In April of 2004, Wizzard acquired MedivoxRx Technologies. Founded in 2000, MedivoxRx is the originator of "Rex", the talking prescription pill bottle, which "talks" to the patient, allowing them to distinguish what type of medication is in the bottle and hear information on dosage and refill instructions www.rxtalks.com www.wizzardsoftware.com IR Contact: Arthur Douglas and Associates, Art Batson, 407/478-1120 Wizzard Contact:Danielle Lewis 412/621-0902 danielle@wizzardsoftware.com

U.S. Army to use MedivoxRx's Talking Pill Bottles in Humanitarian Efforts: MedivoxRx Technologies Inc., a subsidiary of Wizzard Software Corp. (OTCBB: WIZD), announced that it has received the first order for its Talking Pill Bottle from the U.S. Army in Afghanistan. The Army will use the bottles as part of its humanitarian aid efforts to help villagers in Afghanistan properly use the medications that Coalition Forces are dispensing. The Coalition Forces' physicians handle 10,000 children per month in southern Afghanistan's Pashtun tribal region alone. Rex, the Talking Pill Bottle, is the only disposable "talking bottle" now available. Completely self-contained, it can be used anywhere, including the rural areas of Afghanistan. No readers, scanners or playback accessories are required for patients to use it. That makes it perfect for the Coalition Force's efforts to deliver accurate, understandable medical instructions to locals, in their own language and dialect, in an area where people typically are illiterate and rarely speak English. "The first shipment to Afghanistan was made on September 22nd, 2004," said MedivoxRx general manager, Gene Franz. The initial order is for 1,500 bottles to be used in a broad, 30-day field trial. Already successful use of the bottles in informal field trials encouraged the large-scale test. The bottles will be distributed by the Coalition Forces to its medical workers, with each of seven groups receiving about 200 bottles to try in the field. When they leave a village, the army's humanitarian workers typically leave behind a supply of medications with a village elder. A few weeks after the medics leave, the village elder may forget the instructions given for a particular medicine. A simple push on the button at the bottom of the Rex container allows the bottles to "speak" what the bottle contains and how to use it. The recordings are made in the local language, Pashto, or any of the other languages used in Afghanistan. A push on the button and the village elder is able to find and administer the correct medication left by the Coalition Forces and a sick person is on the way to feeling better. The current evaluation of Rex will last approximately one month. At that time, the success of the project will be reviewed. If the large-scale trial concludes as successfully as the earlier informal tests, the Coalition Forces could begin broad deployment of the talking bottles early next year. A humanitarian group also using Rex is the Crippled Children's United Rehabilitation Effort (CURE International), a group based in Lemoyne, PA. The group builds, funds and manages hospitals and supplies surgical, medical and comprehensive rehabilitative care in a teaching context. Its hospitals are networked together, forming a mutual support system that can share personnel, supplies and technology to optimize the effectiveness of each hospital. In addition to its humanitarian uses, the talking bottle has practical applications at home. Overall, pharmacists in the U.S. are expected to fill an estimated three billion prescriptions this year. A national U.S. pharmacy chain might fill 1.5 million prescriptions per day. Roughly two percent of those prescriptions will go to people who could benefit from the use of a talking pill bottle. That is about 30,000 prescriptions per day or one million people per month who can be helped. The usefulness of the product goes beyond making life easier for the blind. Illiterate people, elderly with problems reading fine print, the cognitively impaired and others benefit from hearing, rather than reading, medication information. According to the National Academy on an Aging Society (NAAS), the inability to read prescription pill bottles has profound economic consequences. A NAAS study concluded that the estimated additional health-care expenditures due to low health-literacy skills were approximately $73 billion (in 1998 health-care dollars). The military is already aware of the value of Rex, The Talking Bottle in the States. It is now available in Veterans Administration Medical Centers in locations like Brooklyn, NY and Danville, IL. www.rxtalks.com www.wizzardsoftware.com IR Contact: Arthur Douglas and Associates, Art Batson, 407/478-1120 Wizzard Contact: Danielle Lewis, Marketing Director, 412/621-0902 danielle@wizzardsoftware.com MedivoxRx Contact: Eugene J. Franz, General Manager, 630/682-5910 gene@wizzardsoftware.com

Desktop – Products/Applications/Partnerships/Financials

E-learning : Acapela's speech synthesis makes it easier to teach reading and writing to illiterate adults.
For the international day of literacy, the German government has demonstrated an original initiative : a free e-learning portal aimed at illiterate adults. This web site optimises autonomous learning due to the use of Acapela's speech synthesis. Acapela Group, Europe's leading supplier in speech technologies, has been chosen as the technological partner for the German portal www.ich-will-schreiben-lernen.de (« I want to learn how to write »), which is a free website aimed at training illiterate adults. Only a few hours after the German State Secretary launched the website, on the 8th of September 2004, the portal experienced a great media success with more than 40 000 connections and 1 700 registrations. Germany has around 4 million illiterates, only 0,5% of which follow part time education programmes. The www.ich-will-schreiben-lernen.de website is designed to help all illiterate people: 17 000 interactive courses and various news are available for free, anonymously. Acapela's speech synthesis reads aloud the exercises and the corrections with a pleasant and very natural voice, and enables an easy navigation on the website. Acapela's synthesis is available in 23 languages and perfectly recreates the human voice. It perfectly answers to e-learning and information accessibility needs.. Acapela's speech synthesis is a "text-to-speech" (TTS) solution offering numerous application solutions. These are used by developers using synthesised voices as high quality information supports in their products and services. Based on the 'units selection' technology, Acapela TTS is able to recreate the human voice and can be easily integrated to speech enable any Telecom, Automotive, Multimedia, Mobile devices, Accessibility, Industry and Consumer Electronics applications and services. Acapela Group, major supplier of speech solutions and vocal technologies in Europe, the Middle East and the Nordic markets, is very well known due to its numerous clients and partners in speech projects. These include: Alcatel, Alva, Atos Origin, Bo Jo Tveter, Bosch Blaupunkt, Daimler Chrysler, Franklin Electronic Publishers, Harman Becker, Intervoice, Magneti Marelli, Navigation Technologies, PSA Peugeot Citroën, Telia, Temic, Texas Instruments, Tieman, Toby Churchill, Scandis, Siemens, etc. www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471

Acoustic Magic’s Voice Tracker Array Microphone enables Clallam County to record public Commissioner Meetings and work sessions: Public meetings are often recorded to assure accurate and complete minutes. This presents a challenge to microphone technology, since many of the meetings are held in large, reverberant rooms, where it is difficult to pick up discussion around a table and questions from the audience. Clallam County has met this challenge through the use of the Voice Tracker Array Microphone combined with SoniClear digital recording software from Trio Systems. Public meetings are often recorded to assure accurate and complete minutes. This presents a challenge to microphone technology, since many of the meetings are held in large, reverberant rooms, where it is difficult to pick up discussion around a table and questions from the audience. Clallam County has met this challenge through the use of the Voice Tracker Array Microphone combined with SoniClear digital recording software from Trio Systems. According to Dan Engelbertson, County Administrator, "Our meetings are held in a 40’ by 40’ room in an old Courthouse, with hard walls. There was so much reverberation, it was hard to understand what was recorded. The Voice Tracker is a thousand times better than the conventional microphone we were using. We can understand all the speakers around the table during work sessions, and we can understand comments from the audience 30 feet away" The Voice Tracker locates a talker and electronically steers a "listening beam," like an acoustic searchlight, in that direction. This creates spatial filtering; sounds and reverberation from other parts of the room are not picked up. In addition, digital noise reduction processing removes background noise. This two-stage noise reduction, coupled with the sensitivity of eight microphone elements, gives the Voice Tracker outstanding range and sound quality. Clallam County uses the Voice Tracker microphone with the SoniClear MeetingPro Gov software running on a Windows computer. MeetingPro automatically records from the Voice Tracker microphone and makes it easy for clerks to record meetings, produce minutes, and research action items. The digital recordings can also be archived on CDROM, posted to a network server, and webcast on the Internet, making meeting information accessible to staff, elected officials, and the public. Acoustic Magic, Inc., located near Boston, manufactures the Voice Trackertm desktop microphone that has found application in speech recognition, meeting and transcription recording, VoIP and conventional teleconferencing, voice control for home automation, and security surveillance. The Voice Tracker Array microphone has been Certified by Dragon and ViaVoice speech recognition software, HomeSeer and HAL Home Automation software, as well as SoniClear recording software. www.AcousticMagic.com Contact: Bob Feingold Bob@AcousticMagic.com, or 978 440 9384. Founded in 1978, Trio Systems LLC is a leading provider of recording software and software development components. The company’s product offerings include audio recording and streaming software, and embedded database software components for use in small businesses, local governments, education and retail channels.www.soniclear.com Contact: Alan Bartholomew alan@triosystems.com, or 626-584-9706

Dolphin Supernova Version 6 Language Releases: Supernova version 6.02 will be released this month in Polish. The following languages are now available: *English (UK and US); *Swedish; *Dutch; *German; *French; *Spanish; *Danish; *Czech; *Arabic. Italian, Norwegian and Finnish are expected next month.Customers with v6.01 can download the update for free on the Internet Updater (select Maps menu and Update from Internet or <ALT + M, U>) Language packs will also be available for registered users from our website. The 6.02 update also contains some fixes for customers with Windows 98, and for some known issues with Document Read and Auto Virtual Focus on the Internet. The 6.02 update will also announce when you cursor in and out of the print area in Excel. Contact: Kathy Andrews, Sales, Dolphin Computer Access Ltd Phone: +44 (0)1905 754 577 kathy.andrews@dolphinuk.co.uk Web: www.dolphinuk.co.uk

Oddcast Selects NeoSpeech TTS Voices for VHost Animated Characters and Adds High Quality US English, Japanese, Chinese and Korean to Its Extensive Suite of TTS Languages: NeoSpeech (Fremontm CA) and Oddcast (New York, NY) announced that NeoSpeech's VoiceText Text-to-Speech engine has been integrated with Oddcast's VHost technology. By using VoiceText, Oddcast will be adding high quality US English, Japanese, Chinese and Korean to its existing suite of 10 languages and enhancing its product reach to the Far East. www.neospeech.com Oddcast's VHost software enables companies to improve business results by bringing web and wireless content to life through conversational agents. By integrating NeoSpeech's Text-to-Speech technology, these conversational characters can interact with customers and employees via online advertising campaigns, customer service applications, e-learning systems, and mobile devices, and through many other technology driven platforms. NeoSpeech's VoiceText generates natural-sounding voices from text input and offers the same Text-to-Speech solution across hand-held, desktop, laptop and network applications, while supporting a broad range of industry standard operating systems. Available in several languages, VoiceText can process any size or type of text, in applications ranging from telephone access to information or personal e-mail, to reading complete news stories, delivering the highest level of accuracy and quality. NeoSpeech is a provider of Text-to-Speech (TTS) technology and applications for the mobile, enterprise, entertainment and education markets. NeoSpeech offers flexible, high-quality TTS engines available on the market for U.S. English and major Asian and European languages, supporting a wide range of embedded devices, desktop and network/server applications. Founded in 1999, Oddcast (www.oddcast.com) develops and distributes conversational character software products. Oddcast's core technology, the VHost Platform, enables companies to improve business results by bringing web and wireless content to life through conversational agents. The Oddcast VHost platform is the industry-leading solution for enhancing online communication with intelligent, conversational characters that can deliver information and respond dynamically to customer queries. All of Oddcast's products are Web-based and ASP-hosted and can be delivered to any device even in low-bandwidth environments -- PCs, CD-ROMs and mobile devices. The company serves over 2000 customers, including Coca-Cola, Nextel, Verizon, L'Oreal, Holiday Inn, Washington Mutual, McDonald's and Chevrolet. Contact: NeoSpeech public relations, Jennifer Deitsch, 925-831-2092 jennifer.deitsch@neospeech.com or Oddcast public relations Michael Volpatt, 917-678-0187 michael@larkinvolpatt.com Adi Sideman, CEO of Oddcast.

Harcourt Assessment Acquires Ordinate Corporation: Leading test publisher Harcourt Assessment, Inc., has acquired Ordinate Corporation (Menlo Park, Calif.). The acquisition immediately advances Harcourt’s ability to offer technology-enabled English, Spanish and other spoken language proficiency assessments and testing services to its customers in the United States and globally. It also positions Harcourt for future growth in the field of technology-based assessment. Ordinate Corporation is the first company to develop a completely automated method for evaluating spoken language proficiency based on advanced speech recognition technology. Used by educational, commercial and government organizations worldwide, its tests can be administered anytime and anywhere, and score reports are available in a matter of minutes. The company’s flagship assessment is the Spoken English Test (SET), which is used to assess the listening and speaking skills of non-native English speakers. The 10-minute SET-10 measures language proficiency using five tasks – reading sentences aloud; repeating sentences; answering short questions; building sentences; and answering open-ended questions. SET-10 scores reflect spoken English skill levels in sentence mastery, fluency, vocabulary and pronunciation. Ordinate also offers a shorter, five-minute version of the SET-10, with fewer items and tasks. In addition, the company provides two Junior SET tests for younger learners, as well as a spoken Spanish test. The company is currently developing language proficiency tests in Dutch and Japanese, and it plans to offer additional languages, including French and German. "Ordinate Corporation has been an innovator in spoken language assessment since its inception in 1996," said Aurelio Prifitera, publisher of Harcourt Assessment, Inc., and president of Harcourt Assessment International. Ordinate and its approximately 25 employees will remain in the San Francisco Bay Area. Bernstein will remain as president of Ordinate and Harcourt vice president for advanced language systems. Ordinate’s other co-founder, Brent Townshend, will continue to serve on its advisory board. The Ordinate acquisition is Harcourt Assessment’s third in as many years. Last December it acquired the Dutch testing company Swets Testing International, and in April 2002 it acquired Paris-based Les Editions du Centre de Psychologie Appliquée (ECPA). Ordinate was advised in the transaction by Capital Run LLC. Terms of the acquisition were not disclosed. Harcourt Assessment is a leading provider of high-quality assessment instruments and testing programs. Its educational products include the Stanford Achievement Test series and the Otis-Lennon School Ability Test. Under the PsychCorp? brand, it also publishes a wide variety of clinical products, such as the Wechsler family of psychological tests and the CELF family of speech and language tests. Through the development of custom statewide educational testing programs, Harcourt is helping more than 20 states meet the accountability requirements of the federal No Child Left Behind Act. The company is affiliated with the Harcourt book publishing companies, and together they form the global education brand of Reed Elsevier Group plc (NYSE: RUK and ENL) -- a world-leading publisher and information provider operating in the science and medical, legal, education and business-to-business industry sectors. Privately held Ordinate Corporation was founded in 1996 to develop advanced language testing systems based on new techniques for using speech recognition to score spoken language proficiency. Ordinate combines recognition of spoken utterances with analysis of the manner of speaking to determine language proficiency and other speaker characteristics. Ordinate’s augmented speech analysis technology sets a new standard in language testing for speed, accuracy, cost and convenience. www.HarcourtAssessment.com www.ordinate.com Contact: Ordinate, +1 650 327 4449

Talking, 3D-Animated Character Software for PowerPoint Includes Enhanced Tools for Embedding, Using PowerPoint 2003 Viewer on CD-ROMs: Right Seat Software, Inc. announced the release of Vox Proxy Version 3, bringing new versatility to the process of adding and viewing 3D-animated talking characters in Microsoft PowerPoint presentations. The Vox Proxy collection of fun, effective characters, custom-designed for the PowerPoint platform, can now be viewed using the free PowerPoint 2003 Viewer, while media files can be played using Microsoft Windows Media Player 9. A new online help system has also been added, along with Vox Proxy CD Prep enhancements that allow the PowerPoint 2003 Viewer to be included on Vox Proxy presentation CDs. Vox Proxy, a leading interactive add-in for PowerPoint, uniquely allows users to incorporate talking, animated characters in PowerPoint presentations. Vox Proxy offers intuitive authoring tools to make avatar technology cost effective and easy to use, along with a choice of packaged characters suitable for any type of presentation. Vox Proxy Version 3 gives individuals using Windows Media Player 9 the ability to view media files within Vox Proxy simply by clicking the PlayMedia command. Support is completely transparent; existing files will play on any version of Media Player. CD Prep Add-On To make it easier for people without PowerPoint to view Vox Proxy presentations, Vox Proxy CD Prep Edition Version 3 includes new tools for use when preparing presentation CDs. Creators can specify that the CD should always play on the PowerPoint 2003 viewer, even if the person has PowerPoint, thus ensuring a consistent playback platform. The CD Prep tool also includes Quick Prep, which automatically prepares everything creators of single presentations need for the CD in a single step, as well as an Automatic File Scanning tool for easier file management. New Help System In addition to a number of new commands, macro improvements, and other functional enhancements, Vox Proxy Version 3 offers a new online help system with context-sensitive content, making Vox Proxy easier than ever to use. The help system is also presented in a fully-formatted PDF file for download and printing if desired. Pricing and Availability Upgrades to Version 3 from any previous version of Vox Proxy are available at a price of $99.00 (US). Upgrades for previous users of CD Prep edition are included in the $99.00 upgrade price. For users who do not have CD Prep and wish an upgrade, a special CD Prep upgrade is available for a limited time for an additional $29.95 (normally $49.95). Vox Proxy Standard Edition retails for $199.00 and Vox Proxy CD Prep Edition sells for $228.95. Academic users should call 303/278-2244 or e-mail sales@voxproxy.com for pricing. www.voxproxy.com Vox Proxy sets a new standard for PowerPoint presentations and delivers an innovative, personalized approach to capture audiences' attention. Vox Proxy characters perform numerous animations, speak in multiple languages using either synthetic text-to-speech engines or actual recorded voices, respond to voice commands and carry on interactive conversations. Tom Atkins, President of Right Seat Software, Inc.

Sweden's Sectra Names Vianeta to its Clinical Solutions Network: Sectra - the Swedish IT and medical company - today named Vianeta Communications (Milpitas, Calif.), an official Sectra Clinical Solutions Network (CSN) partner. "Sectra is one of the world's leading PACS (Picture Archiving and Communication Systems) developers," said Ralph Aceves, chief operating officer of Vianeta. The Silicon Valley-based IT firm has pioneered the use of XML technology to create a unified Web-based e-HIM platform called Vianeta Harmony. "Internationally, Sectra PACS are marketed and sold through different partners. Sectra did their due diligence before admitting us into their exclusive network; so this announcement is both a validation of Vianeta engineering and a terrific marketing opportunity on several fronts, including the chance to expand beyond North America." Hospitals and clinics in the U.S. and Canada use Vianeta Harmony - the only native XML health IT solution on the market today - to reduce cost, improve revenue cycles, maximize efficiency, prevent errors, improve quality, ensure patient privacy and build a solid foundation for a true electronic medical record. "We recognize the value Vianeta's speech recognition system offers radiologists," said Sectra Business Development Manager Martin Hakansson. "Both Sectra and Vianeta feature open architecture design so interoperability is never an issue for either of us with any vendor. However, CSN partners like Vianeta integrate clinical applications to Sectra PACS through our open interfaces, to ensure customers the ultimate in successful information exchange. That's what our Clinical Solutions Network is all about: identifying third-party vendors with cutting-edge expertise in specialized fields who can increase the functionality of our PACS products so our customers can relax and enjoy this new technology without wondering if it's going to play nicely with the other systems they also need." By integrating with Sectra PACS, CSN partners provide Sectra customers with the freedom to choose the best clinical application available for solving their particular clinical needs. CSN is also available for university hospitals, giving researchers access to an open platform for developing research applications. Headquartered in Linkoping, Sweden, Sectra is one of today's major diagnostic imaging providers with more than 500 installations worldwide. Sectra's core business is to deliver top-of-the-line, high-availability, robust enterprise diagnostic imaging for 24/7 operation. The market leader in Scandinavia with more than 50 percent of all film-free installations, Sectra also has a large number of systems installed in most major countries in Europe, the Far East and North America, in cooperation with partners. On the Net: Sectra sites: http://www.sectra.com and http://www.sectra.com/medical/partners Vianeta Communications provides XML-based software solutions for health information management that help healthcare organizations of any size or shape automate HIM processes. Hundreds of hospitals and clinics, 60,000 physicians and 4000 HIM staff members throughout North America use Vianeta Harmony products to effectively create, aggregate, manage, process and distribute patient health information, enabling their organizations to reduce cost, improve revenue cycles, maximize efficiency, prevent errors, improve quality, ensure patient privacy and build a solid foundation for a true electronic medical record. Vianeta Harmony combines digital dictation, speech recognition, transcription, document scanning and distribution, and electronic integration with multiple hospital departmental systems, to unify health information into a central e-HIM platform. Clinicians can make care decisions, e-sign documents, and correspond quickly with coders and other internal/external staff by accessing digital patient folders that reside on this unified e-HIM platform from work, home or wherever. Vianeta Harmony significantly reduces chart distribution costs and improves clinician satisfaction with automated printing and faxing to specific locations. Analysts can access chart deficiency information online and track flow of documentation. Coders can access complete patient records online for optimum efficiency thus enabling remote coding. Business offices can fulfill their HIM needs without burdening "Release of Information" staff. And patients can access records and communicate with caregivers online. Incorporating Light Weight Portable Documents, XML and Web technologies, Vianeta Harmony affords 360-degree views and enterprise-wide control of HIM operations, automates workflow to reduce operating costs, and optimizes communication and coordination throughout the enterprise. And unlike other solutions that entail expensive Word deployments and proprietary hardware, Vianeta's open architecture minimizes IT investment and support needs. www.vianeta.com Contact: Vianeta Communications, Carol Somer (408) 519-2156 or (858) 204-7424 (cell) csomer@vianeta.com Sectra, Helena Johansson 46-13-23-52-04 or 46-706-27-52-04 (cell) helena.johansson@sectra.se

Voice Print Unveils Activ! Insight -- Powerful Speech Analytics Technology: Voice Print International (Camarillo, CA) unveiled Activ! Insight -- a speech recognition and call mining application. Activ! Insight provides timely marketplace intelligence that enables organizations to take advantage of emerging opportunities, while reducing operational costs and managing risk. Contact centers can now utilize Activ! Insight in conjunction with Voice Print's full-time recording solutions to capture vast amounts of customer interactions data and automatically derive business insights that business managers need in order to react to the business conditions and opportunities at hand. Powered by CallMiner, Activ! Insight's speech recognition technology converts recorded conversations into text, which can then be searched and analyzed to uncover valuable business intelligence for trending and other purposes. Whereas other call mining technologies available today are often only capable of word spotting, the more intelligently designed Activ! Insight can be deployed to identify the topics within calls -- topic identification yields far more meaningful and accurate information. Activ! Insight facilitates the analysis of sample call records or an entire database of recorded calls, and renders graphical or numerical representations to reveal business trends, opportunities or threats. Activ! Insight listens to calls and returns a digital score on the content of the call -- enabling supervisors to evaluate customer and agent behaviors. "Each day, millions of hours of customer calls are captured and only a fraction of that information is extracted to uncover trends in customer and agent interactions. Activ! Insight provides businesses with real-time intelligence and information that allows them to understand what customers want, and how agents are responding -- a valuable basis for timely business decisions," said Andrew Marsh, CTO of Voice Print International. Organizations traditionally expend massive resources in telemarketing campaigns accompanied by exhaustive playback sessions in order to uncover business intelligence. Activ! Insight takes the complexity out of achieving business insight by focusing on producing an output from the Corporate Information Factory that is highly consumable and actionable. By using Activ! Insight's data mining capabilities, businesses can immediately begin analyzing 100% of their calls. For example, call centers can highlight the number of calls where customers have canceled their accounts or mentioned a competitor's name. Armed with this knowledge, Activ! Insight's analytical tools can be used again to uncover the reasons why an account is typically canceled or what customers find desirable in a competitor. The sudden increase of key words such as "cancel" and "complaint" could prompt notifications and reports to be sent automatically to management to initiate immediate remedial action. Voice Print International is a leading provider and innovator of digital voice recording, screen capture and quality monitoring solutions for contact centers, financial traders, government, and public safety organizations worldwide. www.VoicePrintOnline.com 1-800-200-5430. Contact: Lynn Grogan, 805-389-5200 x5276 LGrogan@VoicePrintOnline.com

Base Technology – Products/Partnerships/Financial

Norwegian Kari joins Acapela High Quality text to speech family Kari, the new high quality text to speech Norwegian female voice, enriches Acapela's multilingual portfolio and answers the Norwegian market's demand for a high end natural voice. Kari can now voice enable any high end applications and services such as unified messaging, email reading, voice portals, e-learning, accessibility tools, public address in railway stations and airports, CRM, virtual agents and speakers, navigation aid and traffic information, etc. Kari's golden tones are likely to make even the most sceptical user sit up and listen... Bo Jo Tveter, a major Norwegian distributor of aids and solutions for the visually impaired has already validated Kari's quality. "Kari is so natural that it is almost impossible to distinguish the difference between a recorded human voice and speech synthesis" said Tore Simensen, Bo Jo Tveter General Manager. Kari already voice enables two key Phoneticom accessibility products, which are distributed by Bo Jo Tveter: ReadSpeaker, a web page reader and Phoneticom Daisy Generator, a publication system allowing the creation of talking books from electronic originals in HTML or XML format. The new Norwegian high quality text to speech can also perform a voice charter through any audio touch point, using Speech Brand. Made possible by Acapela technology, SpeechBrand is a characteristic voice persona used to identify a unique brand or company, in order to build lasting links. A Telecom operator or a car manufacturer can easily speech enable their services with their own, exclusive tone. Acapela High Quality TTS has already been largely deployed in high profile consumer services. With its Norwegian version, Acapela aims to reinforce its leading role in voice-enabled services and interfaces in the Nordic countries, providing innovative and revenue generating solutions. To hear Kari, go to www.acapela-group.com/demos/samples.asp www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471

Acapela Group speaks out in the Middle East: At the Gitex trade show in Dubai, Acapela is speaking out at its stand, but also at those of Speakatel and Etisalat with Salma, its high quality Arabic speech synthesis, unique on the market, that reinforces its leading position in the Middle East. By teaming up with the main specialists in Accessibility and Telecommunications, Acapela Group has become the leader of speech technologies in the Arabic world. Following a long history of partnerships in the accessibility area, Acapela has joined forces with three companies who develop Arabic language screen readers : * Medialog and their screen reader JAWS (with Freedom Scientific) * Nattiq Technologies' screen reader Supernova (with Dolphin) * Visiotech who has developed the screen reader Virgo (with Baum). These active partnerships allow us to offer Arabic-speaking disabled people reliable solutions to gain freer and more independant access to computer technology. In the telecommunications area, Acapela finally gives access to any arabic contents over the phone, with Salma, a very high quality voice. A must for telecom integrators, solution designers, voice platform builders or service providers. And the first results are as good as expected : Speakatel (UAE), Globitel (Jordan), Somac IT (KSA), ATSI (Egypt), Speechworkers (Egypt), Alcatel, Atos Origin, Intervoice, Telia, Siemens, are some of the clients who have already tested, approved and chosen Acapela's Arabic synthesis. To listen to samples of Acapela's high quality Arabic text to speech : www.acapela-group.com/demos/samples.asp www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471

Fonix Adaptive ASR Increases Accuracy, Reduces Error Rates: Fonix Corp. (OTCBB: FNIX), announced a significant technological advance in their automatic speech recognition (ASR) that allows it to adapt or "learn" varying pronunciations and speaker accents. Based on proprietary neural network technology, Fonix ASR improves the accuracy of Fonix speech interfaces without requiring large amounts of memory or processing power. Test results demonstrate significant increases in recognition accuracy utilizing Fonix's new Adaptive Phonetic Optimization. Fonix will immediately begin incorporating this technology into our speech solutions to improve responses to difficult vocabularies and noisy environments. Fonix ASR-based speech solutions for mobile/wireless devices, videogame consoles, telephony products and assistive devices will become available starting Q4 2004. 801-553-6600 and say "Sales" or visit www.fonix.com. Contact: Elizabeth Sweeten, 801-553-6600 mediainfo@fonix.com

LumenVox Announces Version 5.0: LumenVox (San Diego, CA) announced Version 5.0 of their Speech Recognition Engine (SRE), LV Speech Tuner, and Speech Driven Information System (SDIS). The latest version of the SRE offers dramatically improved performance. The main feature implemented in version 5.0 is support of the W3C’s Speech Recognition Grammar Specification (SRGS) and Semantic Interpretation for Speech Recognition (SISR). LumenVox’s Noise Reduction Module (NRM) automatically adapts the SRE to the acoustic environment and dynamically updates its estimate of noise levels. The SRE operates on RedHat Linux 9.0 or later, as well as Windows NT, 2000, and XP. LumenVox allows developers to custom-design their own grammars on-the-fly, filter out-of-vocabulary words, and modify the engine and grammar performance at run-time. Built-in grammars handle single digits, currency, and natural numbers. Version 5.0 of the LV Speech Tuner now allows end-users or value-added resellers to perform tuning, transcription, and testing of any speech driven application created on multiple ASR platforms. This will dramatically expand the usability of the LV Speech Tuner to LumenVox’s partners, who also have deployed applications over other ASRs in the industry. The newest additions to the already feature-rich Speech Driven Information System (SDIS) are full SRGS support and on-the-fly project switching. With this flexible line setting, developers can assign each phone line to a different database, file project or CAPI only mode and then switch to a different project if needed. (877) 977-0707 www.LumenVox.com Contact: ChrisLotspeich@lumenvox.com

Nattiq Technologies Releasing Talks mobile phone reader: Nattiq Technologies L.L.C.(Dubai, UAE) is releasing at GITEX 2004 its state-of-the-art Talks mobile phone reader. For the first time Talks will feature its Arabic Text-To-Speech engine which allows it to read, in Arabic, everything displayed on the screen. As mobile phone units get smaller, their dependence on speech will increase until screens disappear and mobile phones go embedded. Talks is considered a useful tool for many users including busy executives, sales representatives, linguists, school and university students, technology hobbyists, senior citizens, partially sighted and blind users. With a one button touch, Talks announces the caller ID. Users may hear incoming SMS while their eyes are occupied with a more important function. Changing phone settings has never been easier with speech announcement of all setting alternatives. Talks is equipped with Training Mode which enables users to learn the function of all keys without straining their eyes by looking at tinny mobile phone screens. Key echoes help users hear what they type. Talks runs on any of the following mobiles: Nokia 3650, 7650, 6600, 3660, 7610, 9210, 9500, N-Gage and N-Gage QD . It also runs on Siemens SX1 mobile. Speech is available in Arabic and English. Other languages are also available. Nattiq Technologies L.L.C. specializes in speech related products. Founded in 2001 in Colorado – USA, and then moved to Dubai UAE. Nattiq forged strategic alliances with industry leaders such as Dolphin Computer Access - UK, Scansoft - USA, Acapalla Group – Belgium, RDI – Egypt and Ewan Information Technology – UAE. Nattiq helped the localization of state-of-the-art products such as Hal and Supernova screen readers and Talks speech mobile software. www.nattiq.com Contact: Nattiq Technologies +971-4-3326339, Auda Hazeem CEO of Nattiq info@nattiq.com

Nuance Vocalizer 4.0 Brings New Capabilities to Enterprise and Telco Voice Automation Solutions: Nuance (Nasdaq:NUAN) announced Nuance Vocalizer 4.0, the company's newest text-to-speech (TTS) software developed specifically to address the needs of enterprises and telecommunications service providers. With improved voice quality, intelligent address handling and advanced bilingual support, Nuance Vocalizer 4.0 is ideal for applications like directory assistance, account management, voice activated dialing, email reading and others that require the delivery of large volumes of dynamic content via the telephone. Nuance Vocalizer 4.0 also includes several new capabilities aimed at further improving the quality of voice automation applications and simplifying the application development process. Specifically, new prompt-creation capabilities enable companies to replace -- with TTS -- portions of their voice automation systems that previously recorded professional voice recordings. These prompt creation capabilities can further drive down the cost of applications, considering that prompt recordings can cost up to $800 per hour. Nuance continues to offer Custom Voice Creation services for those companies that wish to extend their brands by using the same voice for both their pre-recorded prompts and the TTS portions of their applications for a seamless customer experience. To further meet this need, Nuance will make available in the fourth quarter of 2004 a set of tools to create custom voices, empowering customers and partners to create custom TTS voices on their own, reducing reliance upon Nuance Professional Services or other third parties. These innovative tools leverage Nuance's experience in custom voice creation -- Nuance custom voices will power more than a half-billion calls by the end of 2004. Additional Capabilities for Nuance Vocalizer 4.0 : -- Reduced Memory Utilization: Nuance Vocalizer 4.0 has been optimized to reduce memory requirements by 25% compared to Nuance Vocalizer 3.0, reducing deployment costs. -- Graphical Dictionary and Tuning Tools: Nuance Vocalizer 4.0 includes a dictionary tool that allows words and phrases to be added to the dictionary using a graphical interface. -- Nuance Speak As One: Nuance's Speak As One offering is a set of technologies and services enabling developers to use their chosen TTS voice for application prompts as well, and to tune the transitions between the two. This results in a single, high-quality voice and "personality" throughout the system and a seamless, more conversational and satisfying caller experience. -- Support for Industry Standards: Nuance Vocalizer 4.0 supports the W3C VoiceXML 2.0 and SSML specifications for advanced tuning capabilities. -- Distributed Architecture: Nuance Vocalizer 4.0 leverages the Nuance distributed architecture for fault tolerance, high reliability and scalability. -- Expanded Language Coverage: In addition to U.S. English, Canadian French and Latin American Spanish, Nuance Vocalizer will offer support for Brazilian Portuguese later this year. Nuance supports an additional 13 languages via partners. Nuance partners and other companies that use Nuance Vocalizer software include AAA Minnesota/Iowa, BeVocal, Cisco, Countrywide, Datria, Edify, Gold Systems, HTI Voice Solutions, Interactive Intelligence, IVB, LogicTree, Syntellect, STM, Tellme Networks, TELUS and Workforce Technologies. To experience the sound of the Nuance Vocalizer 4.0 text-to-speech engine, try the online demonstration at: www.nuance.com/prodserv/demo_vocalizer.html Nuance Vocalizer 4.0 is available now. CONTACT: Eastwick Communications for Nuance Becky Quinlan, 650-480-4032 bquinlan@eastwick.com

Create your voice model with Sakrament Personal Voice Master: Sakrament company (Minsk, Belarus), the leading developer in the sphere of Russian speech synthesis and recognition, announces the release of Sakrament Personal Voice Master (PVM) 2.0. The program allows you to create an individual model of your own voice by dictating a set of words. The voice model can be used in other company’s speech synthesis products. This is the first time a user can hear oneself by creating a synthesized voice which reflects individual voice features of every human. With Sakrament Personal Voice Master 2.0 you can create unique voice bases which are as close to the individual human voice as possible. Comparing with previous versions, in Sakrament PVM 2.0 the process of creating an individual voice model is enhanced and simplified. Creating such models expands the spheres where Sakrament speech synthesis engine can be applied. Sakrament PVM 2.0 can be used together with Sakrament TTS Engine 2.0 for reading aloud texts with a user’s own voice, creating voices for interactive voice responders (IVRs), for educational purposes etc. Contact: Sakrament. Alexander Ostapenko, project manager, Sakrament company, tel.+375 172 840650 ales@sakrament.com http://www.sakrament-speech.com/

Portland State snaps up OGI faculty: Portland State University's hiring of 10 computer science professors out of Oregon Health & Science University's OGI School of Science & Engineering will beef up research in PSU's computer science program, which could lead to a better environment for startup companies. at the same time, OGI will be able to redeploy the money it was paying those professors to strengthen the focus of its own computer science department into biological and health applications research. The move benefits both universities, said Jim Hook, recent chairman of OGI's computer science department and one of the 10 new hires at PSU. "We're moving some people around Portland so that they align best with the missions of the two universities," Hook said. "It's a win for Oregon, and a win for both schools." Hook confirmed that some OGI professors could see that with OGI's increasing emphasis on bioscience, biotechnology and other life sciences research, their futures at OGI appeared uncertain. "A lot of people looked at what they were doing, and what OGI's mission was and what PSU's mission was, and said, 'It makes a lot of sense'" to move to PSU. PSU is investing heavily to hire the 10 professors -- about $2 million per year, including salaries and benefits packages. But that $2 million is buying PSU quite a package. The OGI group brings with it $2 million in annual grant funding, 20 Ph.D. students, and a depth of experience in research that has helped win national recognition for OGI's computer science department. Hillsboro-based OGI retains 14 computer science faculty members. OGI also retains its expertise and national standing in the areas of speech recognition, human-computer interactions, adaptive systems and other intersections of computer science with biological and health sciences. OGI will continue to offer degrees in computer science. Funds that supported the 10 departing OGI computer scientists are now available to hire faculty whose areas of expertise align more completely with OGI's shift toward human and ecosystem health, said Ed Thompson, dean of the engineering and computer science school. "This is another step along the strategic path we started on in July 2001," when OGI merged with OHSU, Thompson said. There is no doubt, however, that PSU immediately gains in stature from its acquisition of a group of computer scientists who have a long history of working together and attracting considerable research money. PSU's department grows immediately from 20 faculty to 30. The addition of new faculty to PSU's computer science department is "a huge momentum play for PSU," said David McFeeters-Krone, a Portland-based intellectual property consultant. It can also help create a better environment in which to grow startup companies. The expansion of PSU's research will "draw more bright minds to the region to grow research, and grow companies," said McFeeters-Krone. The sudden expansion of the department "will likely create more opportunities for research, and consequently, more IP," he said. "This IP could then be used to enhance the competitiveness of our current firms, or spawn entirely new ones." Bob Dryden, dean of PSU's Maseeh College of Engineering and Computer Science, agreed that the addition of 10 new faculty "creates a critical mass large enough that we can now grow to national prominence in several specific areas." Greater recognition will generate more research funding, both from government and from corporations, and will also attract "better faculty, post-docs, graduate students and undergrads," Dryden said. While PSU already had a graduate program in computer science -- its Ph.D. program was approved three years ago -- the sudden addition of notable faculty, ongoing funded research projects and graduate students accelerates the normal academic growth cycle by years. PSU's acquisition of some of OGI's computer science faculty also stems a potential loss for Oregon, said Gordon Hoffman, managing partner of Northwest Technology Ventures, which invests in early-stage tech and biotech companies. "We kept these people in the region," said Hoffman, instead of losing them to other institutions. PSU will probably pull in research funding that might not have been available had these particular professors stayed at OGI, said Hook. Because OGI did not have its own undergraduate students, "we've been consistently dinged on our research proposals for not having enough undergrad involvement." Where mergers between different companies can often be contentious, an academic merger is a different animal, said Hook. Furthermore, OGI's and PSU's computer science departments were not competing with one another for grants -- in fact, faculty cooperated on projects, said Cynthia Brown, chair of PSU's computer science department. Two professors have had long-standing dual appointments at the two institutions. In the end, the change is "a very good deal for the region" with respect to economic development and the growing status of both OGI and PSU, said venture investor Hoffman, who has worked with both OGI and PSU. Hoffman headed OGI's technology transfer office before the merger with OHSU, and now serves on the board of PSU's foundation. "Strong universities and a good startup environment go together," he said.

Phonetic Systems' Voice Search Engine 5.8 Breaks Deployability Barriers in Carrier, Call Center and Enterprise Speech Applications: Phonetic Systems announced a major upgrade to its patented speech recognition technology which is the core foundation of its Enterprise, Call Center and Carrier speech applications. Version 5.8 of the Voice Search Engine (VSE) is an open and standards based speech recognition engine which excels at powering applications requiring large and dynamic grammars. It includes a number of product enhancements such as industry leading performance analysis tools, platform robustness, improved tools for database normalization, and improved multi-modal support. With VSE 5.8, companies can accelerate time to market, simplify their application maintenance, lower costs and increase performance. Executing on its strategy to provide carriers with speech automated public Directory Assistance (DA) and enhanced services, the company has been awarded more than eight carrier DA contracts, allowing them to leap ahead of the competition in terms of percentage of deals won. This significant achievement highlights the company's commitment and investment in developing innovative technology and their escalating momentum in the speech industry. The company has also dominated large scale enterprise and call center directory applications with hundreds of leading companies choosing to use Phonetic Systems' VSE to deliver large scale directory-based speech solutions--such as automated attendant, address change, and portal applications--that require little maintenance. One of the principal motives for customers choosing Phonetic Systems is that applications work out of the box without any manual tuning of grammars in order to keep applications working at optimum levels and therefore delivers a better ROI model. Phonetic Systems' success and extensive experience in large scale deployments cannot be underestimated. The expanded capabilities of VSE 5.8 are a direct result of working with customers (like AT&T, Bank of NY, Procter & Gamble, Fonecta, Pelephone and others) to determine how to better improve the technology. Like previous versions, V5.8 delivers solutions that are quick to deploy, easy to maintain, provide a quick ROI and that deliver more flexible features for customers to better leverage their investments in speech. Customers report record-breaking paybacks in as little as six months. VSE 5.8 offers the following features: Tools, processes and features have been added into VSE 5.8 to support full automated National DA for businesses, government, residential and reverse DA listings. In the area of database performance, enhancements have been made in aliasing, normalization and rule generation. For database editing, additional tools for sorting and updating a GUI are now available. New tools have also been developed for expanded locality creation and NDA database preparation, update and distribution and session log analysis. These lead to overall ease in supporting large geographically diverse data sets. VSE 5.8 further streamlines maintenance of applications and provides the essential features of a managed service for the price of a conventional service. New automated processes for professional recordings and system analysis keep customers' applications running smoothly. For automated attendant and operator services applications, there is now the opportunity to choose between having a male or female voice answering a company's calls allowing for improved customization and flexibility. New features and tools in VSE 5.8 make speech applications even easier to deploy for partners. Some of the enhancements include: Open API for better ability to customize and achieve greater levels of redundancy and scalability, new testing tools for testing applications with other 3rd party platforms, multimodal support for both voice and text searches, and XML support. Partners that benefit from VSE 5.8 include Envox, Voice Genie, IBM, V-Enable, Volt Delta, Telcordia, Soleo, XTEND, DCC, and others. Additionally, VSE 5.8 includes the following features and capabilities which benefit all customers and partners: -- Performance Analysis Tools - The VSE 5.8 performance analysis process boasts the best in the industry. New tools deliver improved automated analysis and logging tools for ensuring superior levels of accuracy. Automated processes monitor and analyze a customer's system performance allowing for proactive maintenance and improvements. -- Architecture - VSE 5.8 offers increased reliability, scalability and the robustness. A modular, distributed approach sits like an ecosystem and automates the transfer of voice and data between the database, switch and workstation. It will run on either Windows OS or LINUX Redhat 9.0 and supports VXML 2.0, SNMP and SNMP viewer. -- Improved Speech Recognition and Accuracy - There is increased accuracy in quickly identifying intended requests and completing calls regardless of connection, accent, and environmental factors from analog, digital, wireless and/or VoIP interfaces. VSE 5.8 has improved proactive data tuning and filtering to boost search performance and bulls-eye processing is accurate enough to not require confirmation. Digit recognition is improved with better recognition of digit pairs and triple digits. This is helpful in recognizing numbers for directory assistance, account numbers and other number lookups. The VSE can function as both a phoneme decoder and a small grammar decoder. As a small grammar decoder, it will get a grammar in GRXML format along with semantic tagging and create an HMM network. -- New Language Development - VSE 5.8 dramatically improves the rapidity to which new languages can be developed. A new language can be developed in as little as six weeks. VSE 5.8 currently supports English, French, German, Spanish, Finnish, Dutch, French Canadian, French Canadian English, UK English, US Spanish, Hebrew and additional languages are currently in development. Through participation in the SpeechDat Across Latin America (SALA) II program, additional South American will be developed. Phonetic Systems delivers real-time speech recognition solutions that use the power of voice to increase productivity and improve customer service. Customers and partners worldwide use Phonetic Systems' Voice Search Engine (VSE) to automate a range of applications like carrier directory assistance, voice access to CRM applications, password management, automated attendant and more. www.PhoneticSystems.com. Phonetic Systems Heather Howland, 978-439-3600 hhowland@phoneticsystems.com

ScanSoft Commences Search for Independent Accounting Firm: ScanSoft, Inc. (Nasdaq: SSFT), announced that its audit committee has commenced a search for a new independent registered public accountant, including requesting proposals from other accounting firms, following PricewaterhouseCoopers LLP's ("PwC") resignation as the company's accounting firm. PwC's reports on ScanSoft's consolidated financial statements for the years ended December 31, 2002 and 2003 did not contain any adverse opinion, or disclaimer of opinion, nor were they qualified or modified as to uncertainty, audit scope or accounting principles. The company noted that there have been no disagreements with PwC on any matter of accounting principles or practices, financial statement disclosure or auditing scope or procedure. ScanSoft does not expect any material adjustments to its previously reported financials as a result of this announcement. www.scansoft.com

VoiceXML Forum Launches Platform Certification Program; Successfully Completes Trial Invites Vendors to Submit Platforms for Independent Testing: The VoiceXML Forum announced the availability of the VoiceXML Platform Certification Program. The Program consists of a series of tests to determine whether a VoiceXML platform is fully compliant with the latest release of the VoiceXML standard. Companies wishing to have their platforms tested may register on the Forum's Web site at www.voicexml.org, or contact the VoiceXML Forum at voicexml-admin@voicexml.org. VoiceXML-certified platforms will be listed on the Forum's Web site. The general availability of this program follows the completion of a successful pilot program conducted with CT Labs, an independent certification laboratory contracted by the Forum, and member companies Nuance (www.nuance.com), VoiceGenie Technologies, Inc. (www.voicegenie.com) and Voxpilot Ltd. (www.voxpilot.com). As in the pilot, the certification testing will be conducted by independent testing organizations chosen by the Forum. The VoiceXML Platform Certification Program test suite includes over more than 600 tests that cover the W3C VoiceXML 2.0 Recommendation and is based on the W3C's VoiceXML 2.0 Implementation Report. To aid vendors in preparing for certification, the Conformance Test Suite to be used in the VoiceXML Platform Certification Program is available, at no charge, for public download from the Forum's Web site at www.voicexml.org. Forum Sponsor and Promoter members also can download a Web-based test harness to aid in running the tests. The VoiceXML Forum Conformance Committee manages the Platform Certification Program with support from the IEEE-ISTO (www.ieee-isto.org) and NCC Group (www.nccglobal.com). The Conformance Committee is currently at work on a program to test VoiceXML applications and tools for conformance to the standard. www.voicexml.org. The VoiceXML Forum is an industry organization formed to create and promote the Voice Extensible Markup Language (VoiceXML). With the backing and contributions of its diverse membership, including key industry leaders, since its inception, the VoiceXML Forum has successfully driven market acceptance of VoiceXML through a wide array of speech-enabled applications. Currently, the Forum has more than 325 member companies, distributed across three membership categories (Sponsor, Promoter and Supporter). Membership in the Forum is open to any interested company. The core activities of the VoiceXML Forum are driven by its five active committees: Accessibility, Conformance, Education, Marketing and Tools. The Forum publishes the VoiceXML Review, an electronic magazine dedicated to VoiceXML, featuring insightful articles written by industry experts. Its public Web site, www.voicexml.org, is a rich, one-stop source for information about VoiceXML and related technologies. www.voicexml.org or contact the VoiceXML Forum Administrative Team at +1 732 465 6464 or via e-mail at voicexml-admin@voicexml.org. The VoiceXML Forum is organized as a program of the IEEE Industry Standards and Technology Organization (IEEE-ISTO). Contact: Prequent, Inc. (for the VoiceXML Forum)
Christian Danella, 408-307-1236 christian@prequent.com or For all inquiries please contact:
VoiceXML Forum Cindy Tiritilli, 732-465-6464 voicexml-admin@voicexml.org

Industry People

VeCommerce has named Steven Fitzjohn and Philip Clipperton as senior business development managers. Fitzjohn will look after the southern region and Clipperton the northern region. TuVox announced that it has named Azita Martin to the position of vice president of marketing. Nuasis Corporation announces that Joseph McFadden has been promoted to vice president of Corporate Marketing and Product Management. ScanSoft, Inc. (Nasdaq: SSFT) announced that it has named James ("Jamie") R. Arnold, Jr. to the position of senior vice president and chief financial officer. Acapela Sweden has welcomed two new members to enhance support and service delivered to its customers and partners. Maria Larsson will be your contact for Financial issues and Zoltan Nagypal will take care of Customer support and Quality process. HTI Voice Solutions, Inc has hired Andrea Krumm as the Regional Sales Representative for Southwest, USA and Andrew Passo as Sales Representative for the Mid-Atlantic, USA.

Financials

Table F1

Q2, 2004 Financial Results

Supplier	Item	Q2, 2004			Q2, 2003	Q1, 2004
		$M	2003 to 2004 % change	Q1, 2004 to Q2, 2004 % change	$M	$M
Andrea Electronics	Total Revenue	1.29	3.7	(17.3)	1.24	1.56
	Net Income	(0.42)	136.9	20.4	1.13	(0.52)
	Cash or equiv.	2.38	(29.1)	32.0	3.35	1.80
Captaris	Total Revenue	18.19	0.3	0.1	18.14	18.18
	Net Income	(0.06)	-	(380.0)	0.71	0.02
	Cash or equiv.	58.67	61.9	2.6	36.24	57.19
Comverse	Total Revenue	233.40	23.8	5.4	188.50	221.40
	Net Income	13.30	1,309.1	(90.0)	(1.10)	7.00
	Cash or equiv.	2,186.76	20.9	1.9	1,808.90	2,146.81
Enghouse	Total Revenue	16.50	29.9	49.6	12.70	11.03
	Net Income	3.20	6.7	44.6	3.00	2.21
	Cash or equiv.	85.66	20.8	37.0	70.91	62.55
Intervoice	Total Revenue	44.30	6.5	1.8	41.60	43.50
	Net Income	5.10	41.7	59.4	3.60	3.20
	Cash or equiv.	49.00	76.3	19.8	27.80	40.90
Preferred Voice	Total Revenue	0.31	(34.0)	(8.4)	0.47	0.34
	Net Income	(0.09)	(25.7)	(493.7)	(0.12)	0.02
	Cash or equiv.	0.94	2,238.6	1,346.4	0.04	0.06
SafLink	Total Revenue	0.92	449.7	14.5	0.17	0.80
	Net Income	(2.39)	(29.2)	20.4	(3.37)	(1.98)
	Cash or equiv.	8.18	(32.9)	(34.1)	12.18	12.40
Verint	Total Revenue	60.17	28.3	6.3	46.89	56.60
	Net Income	5.67	42.6	274.6	3.98	1.51
	Cash or equiv.	228.71	70.8	9.7	133.93	208.50

 

Table F2

Speech Industry Financial Summary (October 1, 2004)

Company	Symbol	Shares			2003		Market Cap.		Mkt Cap./Rev.	2002
		Last	52-wk.		Rev.	Net Income	9/1/04	% change		Rev.
			Hi	Low	$M					$M
Andrea Electronics	AND	0.09	0.90	0.09	5.1	(4.6)	4.78 M	(9.81)	0.93	7.2
Apropos Technologies	APRS	3.00	6.61	2.30	20.1	(5.6)	51.92M	(4.38)	2.58	20.9
Artisoft	ASFTC	2.75	4.65	1.65	6.6	(7.1)	10.51M	45.97	1.59	4.5
Aspect Communications	ASPT	10.42	19.95	7.33	363.8	(29.0)	0.617B	22.66	0.00	396.1
Brooktrout Technology	BRKT	8.96	23.18	7.10	74.7	0.6	116.5M	4.20	1.56	73.5
Captaris	CAPA	4.31	7.50	4.15	83.3	12.5	136.0M	(12.37)	1.63	94.6
Cognitronics	CGN	3.30	4.50	2.15	10.3	(5.4)	18.87M	(5.08)	1.84	11.3
Comverse Technology	CMVT	19.80	20.87	15.25	765.9	(5.4)	3.88B	14.79	5.07	735.9
Convergys	CVG	13.64	20.80	12.30	2,288.8	171.6	1.94B	(1.52)	0.85	2,286.2
Covalent Group	CVGR	2.99	4.31	2.11	20.8	(0.6)	39.84M	13.83	1.91	29.2
Eckoh (UK)	ECK	8.00	17.70	6.00	104.8	(16.4)	39.5	(4.48)	0.38	104.8
Enghouse(Canada)	ESL	9.25	10.00	6.00	34.6	7.2	173.9	2.21	5.02	13.9
Fonix	FINX	0.250	1.59	0.10	2.4	(13.5)	23.57M	67.88	9.89	3.1
Glenayre	GEMS	1.78	4.30	1.43	58.2	1.6	118.68M	(2.14)	2.04	67.4
Interactive Intelligence	ININ	3.53	6.70	2.72	51.2	(5.9)	56.03M	(7.05)	1.10	47.8
Intervoice	INTV	11.49	18.00	7.23	156.2	(66.4)	413.4M	37.49	2.65	145.6
iVoice	IVOC	0.0008	0.038	0.0001	0.4	(2.0)	5.2	(27.27)	11.77	0.6
NMS Communications	NMSS	5.25	9.35	2.41	87.1	(36.0)	245.15M	15.03	2.81	76.5
Nuance Communications	NUAN	4.38	9.40	3.53	55.0	(19.3)	155.62M	10.36	2.83	44.1
Preferred Voice	PFVI	0.130	0.33	0.05	2.1	(1.3)	3.85M	0.26	1.81	2.2
Prosodie (France)	PRD	21.10	28.21	14.80	203.5	(7.5)	189.6	(8.20)	0.93	203.5
SAFLINK	SFLK	2.66	4.90	1.98	2.0	(10.7)	208.7M	30.03	105.40	2.0
ScanSoft	SSFT	4.18	6.50	3.61	135.4	(5.5)	441.54M	(3.81)	3.26	106.1
TALX	TALX	23.55	28.78	18.50	126.1	13.0	323.46M	8.26	2.56	137.8
VoxWare	VOXW	0.035	0.180	0.030	4.6	(6.6)	19.32M	6.39	4.20	4.6
Verint	VRNT	37.83	38.11	18.84	192.9	17.9	1.17B	19.27	0.01	131.2
Voiceware (ROK)		1.66	3.37	0.88	3.8	(1.6)	18.8	(2.93)	4.90	3.8
West Teleservices	WSTC	29.84	29.95	20.30	988.3	88.0	2.01B	18.24	2.03	820.2
Witness Systems	WITS	16.96	17.41	4.87	108.04	(20.6)	397.85M	26.09	3.68	67.69
Wizzard Software	WIZD	1.84	4.39	0.38	0.470	(2.0)	46.94M	12.86	99.77	0.346
Z-Tel	ZTEL	0.50	4.79	0.29	289.1	(16.1)	18.74M	(1.94)	0.06	235.3
Total	Total				6,245.9	23.3	12,895.3	11.80	2.06	5,877.9

 

 

 

 

 

CALENDAR

Subscription Order Form

Please enter/renew my subscription to ASRNews for one year (12 months).

Trademarks mentioned in this publication are the property of the companies mentioned; they are used editorially. The material and statements herein are based on data from sources believed to be reliable, but is not guaranteed as to accuracy and does not, unless otherwise stated, purport to be complete. VIA disclaims any obligation to subsequently update the material or statements contained herein.

ASRNews is published monthly by Voice Information Associates, Inc., P.O. Box 253, Concord, MA, 01742, USA, Voice. 978- 266-1966, Fax: 978-263-3461, http://www.asrnews.com, email, asrnews@asrnews.com Walt Tetschner, Publisher and Editor. Subscription, annual: hard copy only, U.S. and Canada, $345, foreign, $395, Copyright 2001, ISSN 1051-4163. All rights reserved. Reproduction or photocopying is strictly prohibited without written permission. Voice Information Associates is an information services company specializing in the advanced speech technology industry.