ASR News Issue Volume 15 No. 12 December, 2004

How well do we use it ourselves? This month we called the main telephone number of the North American speech companies to determine how they were using speech in their business. For 122 companies, we recorded how they responded to a caller that called the company. This was a followup to a study that we did in 2003. Out of 122 companies, 57 (46.1%) of them utilized speech as an automated attendant. Seven them are live, which left 58 of them with a Touch-Tone auto attendant. In addition to the speech vendors, 6 media companies that cover the speech segment were included. All of these utilized T-T or live operators and not speech.

Many implementations abuse the caller: Most implementations are enterprise-centric, not caller-centric, as they should be. For both speech-enabled and T-T implementations, greeting lengths and the length of the instructions were quite long. The speech-enabled implementations had average instructions that were approximately half as long as the T-T implementations, but were still excessively long. Automated call handling systems typically suffer from unnecessarily long greetings and instructions. Most of them could be improved dramatically by simply not saying anything that a live agent would not say. Greeting needs to identify the company that is being called. Nothing other than this is really needed. A holiday greeting or a pleasant good morning is okay, but recognize that you are taking the callers time. A company hype (the leading,....) is only irritating to the caller. You get music and strange sounds (I accidently dialed a porn line?). The department routing option (for accounting, press or say ) is something that most companies insist on including in their long-winded instructions to the caller. They accomplish nothing other than irritating the caller/customer. My experience is that if I select one of these departments, I get either voice mail or an assistant that promises to pass on a message and get back to me. I invariably do not get anyone to call me back. It's pretty obvious that this entire process is a waste of the callers time. The better approach if you aren't able to connect the caller to the person that they request, is to quickly get them into voice mail. The speech systems having shorter caller instructions seems good until you realize that most of them have multi-stage error recovery schemes that go on over-and-over. Variants on did you mean ____ . As a reference point, live operators are much more consistent. They invariably state the name of the organization and often say good morning. Live operators must have a strong standards organization or perhaps they just are using common sense. The most fundamental problem with most of the automated attendants was that they were attempting to accomplish too many things with a single phone line. This resulted in presenting numerous options to the caller. The caller, that is ready to say the name of the person that they are trying to reach and is not in a listening mode, gets a menu dumped on him/her. We then expect the caller to respond properly. Terrible human factors. We simply are not meeting the needs of the caller. A massive improvement to all of the systems could be accomplished by simply having a mature & seasoned live operator record greeting/instructions. If the live operator wouldn't feel comfortable saying them, then the automated system should not say them. Many of our systems are tied to industry heritage. I still encounter voice mail box messages in which the recipient has decided to go on a long spiel to let me know precisely why he can't take my message in real time. I'm having my time wasted and I really don't care why he isn’t around. Of the 122 systems that were tested, 26 of them did provide a mechanism for capturing the call if the caller did not pass the test of the enterprise being called. They simply hung up if the caller did not respond properly. Imagine that! Over 21% of the speech industry companies did not have the good sense to assure that a call from a potential customer was captured.

If we followed the basic rules shown above, we would have significantly more satisfied callers and more calls would be automated. If we meet the expectations of the caller, speech self-service systems will work much better. Let the caller that is ready to say the name of the person that they are trying reach , say it and get through. Don’t waste the callers time by dumping directions/options at him/her that are not going to be heard.

How about regulating the VUI? As a result of the outcry re outbound calling, we now have strict regulations that attempt to control this. We have a Do-not-Call list that 60 million Americans have signed up for. Regulations governing the operation of predictive dialers have been in place for over a year. The Third Party Validation (TPV) industry was created by FCC regulations. This occurred because the industry was not able to regulate itself. A similar situation is developing for inbound call automation. Few people have a good opinion of call automation systems. They frequently offend and abuse the caller. Virtually everyone has their favorite horror story. Every week the media bashes call automation. I have nine grandchildren and I avoid telling them what I do. The call automation industry has earned a very negative reputation and calls for regulation are a distinct possibility. Despite this being recognized as a major telephone speech industry problem, virtually nothing is being done to really improve this. The media/analyst community continues to function as cheerleaders. A number of independent groups are testing systems and publishing the results. This is fine, but we already know the VUIs are bad. Fixing them is certainly a good thing to do, but this is only a bandaide. The industry needs processes and standards that reduce the number of bad implementations that occur. I’m not yet advocating this but, given the head-in-the-sand orientation of the speech industry, regulation may be the only way that the speech industry is going to come to grips with this basic problem.

Another approach to Name & Address Capture : I ran into a chart at: www.conservit.com/news/2004/11/name-address-application-natural.html that I thought was interesting. It compares the performance of live agents to ASR to a technology pioneered by ConServIT that they call Natural Speech Technology (NST).

When you look at Table E2, the bottomline cost figures suggest that the Apptera ASR and the ConServIT NST approaches are comparable. They both yield a cost per N&A that is under $1 which puts each significantly below the $6.25 per N&A that a customer service agent would cost. Admittedly, for the Apptera ASR approach, for two-thirds of the calls, the N&A is not captured, but for the ones that are captured, the two approaches have comparable cost profiles. When you consider the impact of angry customers, the picture changes dramatically. The cost that is not shown in this table is the cost associated with really irritating a customer. For the Apptera ASR approach, we have jerked the caller around with the "did you mean,,,,,," stuff for a relatively long time already. For two-thirds of them, we were not successful in capturing the N&A. This means that we now can use a customer service agent to do it (at a much higher cost) or we have the opportunity to screw-over the customer once more by subjecting him to the ASR system again. An angry customer is a real cost liability. They are easily converted into ex-customers. It doesn’t take many of these before you have lost all of the cost-saving-ROI that the speech-enabled system provided. For many N&A automation applications, the ConServIT NST is a better solution than speech-enabled ones. In fact, from the data presented in the table, the ConServIT NST is actually superior to a live agent. It doesn’t irritate customers and still saves more money than speech-enabled implementations. www.ConServIT.com Peter Theis, 800-994-4400, Theis@ConServIT.com

More bashing of speech systems: The following appeared in the December 31, 2004 edition in the Asbury Park Press as a recommended resolution to make for 2005: "For companies with voice-recognition systems: Stop using them. Immediately. They don't work and they are more annoying than sand fleas at the beach. Talking to a computer is an indignity. Talking to one that doesn't understand the word "yes" or "no" is the ultimate indignity."

Sensory's New Technology Launch: The announcement that Sensory is making at the Consumer Electronics Show (CES) is the result of a unique investment that Sensory made in 2000 and four years of subsequent hard work. In October, 2000, Sensory acquired Fluent Speech Technologies, a speech technology company founded by researchers from the prestigious Oregon Graduate Institute. The acquisition was targeted to help Sensory expand its low-cost, high-quality speech recognition technology line beyond its current areas of strength - embedding speech recognition capabilities into consumer products such as personal electronics, toys, automotive products and telephone handsets - to new markets, such as PDAs, Internet appliances and higher-end telephony and automotive applications. The results of that investment are that Sensory is simultaneously able to strengthen their unique position as a supplier of both ICs and speech technology for the low-end consumer segment and also enter the high end cellular phone/automotive market segment with a highly competitive product. Very impressive to see something like this! The management of Sensory set a target to accomplish something and actually did it. In this instance, they were betting the future of the company on this effort. How refreshing to see a company forecast what they are going to do and then actually do it. Will be interesting to see how well Sensory does in the high-end speech embedded market segment. Given their track record, I would not bet against them.

Contact Solutions doing well: Contact Solutions is a call automation company that appears to be getting a good deal of traction. Given the tepid overall status of the telephone speech automation industry that we discussed in our last issue, this makes them quite different than many speech technology companies who are struggling. I can identify three things that appear to solidly set Contact Solutions apart from the rest of speech industry. Firstly, the fact that Contact Solutions views itself primarily as a call automation company and not particularly a speech technology company, has something to do with their success. The focus then becomes the caller and how best to meet the needs of the caller. The caller experience is the guiding target and speech becomes nothing more than one among multiple contributing items. Providing personas is not a worthwhile endeavor if you do it at the expense of providing C-TI. The second item is that Contact Solutions appears to fully understand that enterprise data is not all nicely tucked in IP structures (despite the wishful thinking of the VoiceXML folks) and this is not likely to change significantly in the near future. Finally, Contact Solutions appears to be oriented towards truly new call automation applications that harness call information in real time in order to make it actionable. The recent success of the Voter Line Alert application that Contact Solutions ran, has the potential for spawning a broad range of applications that are truly new and revolutionary. In the 20 years that I have been working in the call automation area, this is one of the first applications that I would identify as being really new. Virtually every other one that I’ve seen that claims to be new turns out to be simply a rehash or upgrade of what has been done for the last 20 years. Obtaining data in real-time that is also actionable in real-time has an application scope well beyond the voting irregularity monitoring/reporting that was run. The response of callers to advertisements, weather changes; catastrophes, etc., can be obtained and acted upon in real-time. If a particularly advertising message isn’t working, you can adjust it in real time to obtain a desired response. My betting is that Contact Solutions will continue to be successful in the call automation area. Smart people with a strong experience in the call automation, taking a pragmatic and rational approach.

Bad Ideas Never Die: Ever notice how bad ideas never completely die? I’ve found that this is usually because folks don’t take the time to really understand why it was a bad idea and failed in the first place. Talking Yellow Pages (TYPs) were a hot item in the mid-1980s. Heavy-duty investments were made in this area. Rather than letting your fingers thumb through the yellow pages book, you could call a number, interact and get connected to whatever you were looking for. The basic flaw with TYPs was that this was a solution looking for a problem. The Yellow Pages phone book worked pretty well. 15 years later TYPs reappeared with new names such as voice-web and voice portal. These died also for the same basic reason – a solution that was looking for a problem. Like the Phoenix rising from the ashes, it’s predictable that the concept of a portal that is accessible by the telephone will be "discovered" again. The proponents refuse to accept that it was simply a bad idea. They will claim that it was never bad, just premature and perhaps over-stated, and anyway, improvements in technology will change things. This is nonsense! A bad idea does not improve by throwing additional technology at it. Let it die! Focus energy on the good ideas.

Another bad idea: The concept of doing automatic translation from spoken conversation in one to language to another language is rising as another idea that is being resurrected from the ashes. This was a favorite of the L&H groupies. Take three technically imperfect technologies (ASR, Machine Translation (MT) and TTS, combine them, and what would you expect. Rubbish of course! The basic premise that the researchers are putting forth is that we have not dealt with the context in which something is said and that we need to do this. We need to figure out what the speaker was thinking and not just focus on the words that are said. Although you could rationalize this as an R&D activity, considering that we have not really started to apply the basic human-machine techniques to speech recognition or TTS, it seems to be foolish to somehow skip over the basic steps of dealing with ASR, MT and TTS individually before mushing them all together.

One-Number Concept: Another bad idea. Before DNIS was widely available as telephone network capability, some validity existed for doing this. Today, it’s simply another bad idea that abuses the caller. Attempting to put a variety of different functions on a single phone number usually does not make any sense at all. I recently had to report a power outage to my utility company (N-Star) and call AAA to have my car towed. In both instances I was confronted with a menu that would have permitted me to do a variety of things that I had absolutely no interest in. When my power has gone down, I really am not in a mood to deal with listening to 7 menu choices that include billing inquiries, gas leak reporting and no heat reporting among the choices. When I call AAA to have my car towed, I’m really not terribly interested in finding out how to become a member or obtain maps or other travel information. Implementing self-service systems this way seems to be coming from some sort of mis-guided notion that this is what the callers want. A single number that they can call to reach an enterprise seems like a nice thing until you get into the real world. Enterprises invariably print separate numbers for each of the major functions. Interestingly, neither one of these are speech-enabled services. N-Star self-service is one of these "press or say 1" things, that provides little of the potential benefits of a speech-enabled implementation. If this were an isolated instance, it would be one thing. Unfortunately, this is common. When I order prescriptions, I’m slammed with options that let me check the status of my order or expedite its delivery, although I’ve never had an occasion where I ever desired any of these options. Enterprises just can’t seem to resist rattling off a long list of options to a caller, even when they know the caller is calling only about a single specific thing. Just another example of the VUI designer not understanding the needs of the caller.

A good idea – Speech is about improving customer service: Most customers don’t really believe the cost-saving ROI story that they hear from the speech sales person. If they really believed it, then there wouldn’t be a 9 month sales cycles. Cost-saving-ROI arguments have been put together by MBA-types, not call center operations people. Claims of large cost-saving-ROI invariably disappear when you ask to see data to support the cost savings that were achieved. Requested this a number of times and never received anything. Reason is simple – cost-saving-ROI is a myth created by speech industry marketing folks. Customers accept it because the speech industry has told them that this is what speech is all about. This is one of the dumber speech industry ideas and should be totally replaced by the message that speech is about improving customer service. The argument that most enterprises have little interest in improving customer satisfaction is probably a valid one. Reason for this is simple – the speech industry has chosen to tell them that speech is about cost-saving-ROI. The speech industry needs to tell customers that speech is about improving customer service. That is a really good idea.

The customer made me do it! I hear this over-and-over. The customer playing an active role in the design of the self-service system is identified as the biggest factor causing lousy VUIs. It also is identified as a large contributor to the huge cost of implementing VUIs. The telephone speech self-service area is not totally unique in having customers play a huge role in the design process, but it appears to be much worse. Have you ever seen a neurosurgeon that invites members of the family to assist with the surgery? Admittedly, you can argue that this is a matter of life or death. On the other hand, we’ve all seen the signs for craft services:

These are quite effective. Customers get the message and let the plumber fix the toilet without their assistance. This suggests that your plumber has better business/marketing skills than the leading speech industry vendors have. The speech industry has managed to simultaneously promote the message that a speech implementation is an extremely complex endeavor that requires the skills of a broad set of specialized and highly trained technologists and that individuals within the customer’s organization (that invariably know almost nothing about the needs of the callers (customers)) can also contribute in a useful fashion. Just another example which illustrates how fundamentally flawed the VUI design activity is. The speech industry needs to get the customer, who is generally unqualified, out of the VUI design process.

Call Centers don’t know if their callers are satisfied: In a recent issue, Call Center Magazine published the results of a study on customer satisfaction. Customers: Not As Happy As You Think (www.callcentermagazine.com/shared/article/showArticle.jhtml?articleId=42900017&classroom=) They interviewed both call center managers and callers. From the data that they obtained, it appears as though a major disconnect exists between how satisfied the caller actually is and how satisfied the caller center manager thinks they are. "Sixty-five per cent of outsourcers and 56% of call center workers think their customers go away "highly satisfied" after a call. Unfortunately, only 22% of the customers said that they were highly satisfied. Call centers think that only 2% of their callers go away "not very satisfied." The callers say other wise: a stunning 29% of consumers and 36% of business callers are not satisfied. Certainly room to debate why this is so, but one conclusion is difficult to avoid: Call centers do not know how satisfied or unsatisfied their callers are and somehow believe that they are much more satisfied than they really are.

Do call center managers care about customer satisfaction? I recently had a senior VUI implementation manager tell me that customer satisfaction is not something that he has seen call center managers managers identify as important. This person has been involved with as many speech self-service implementations as anyone in the industry and I have a lot of regard for his opinions/views. Not at all surprised to hear this. The leading speech industry vendors have told the customers that cost-saving-ROI is what speech is all about. When you constantly talk about saving money, customers reach the conclusion that this is what it is all about. The customers simply believed what the speech indsuty vendors were tellingthem. Why shouldn’t they? If the speech industry vendors don’t know best what their product is for, who does? Call centers simply do not associate speech self-service with improving customer satisfaction because this is not a message that they have received from the vendors.. Does this mean that call centers don’t think that customer satisfaction is important? Of course not! Look at the quality monitoring segment of the call center. This area is about training agents to do a better job and satisfying callers is at the top as a success metric. Cost-saving-ROI is not part of equation. It’s also an area that is exhibiting solid growth. The leading vendor in the quality monitoring segment, Witness Systems, is doing a lot better than the leading speech technology vendors. Growth profile is much better. Witness is profitable. Stock price has almost doubled in the last year. Compare this to Nuance and ScanSoft: that are not profitable; have had little organic growth and have stock prices that have dropped significantly during the last year. Perhaps the speech industry vendors should look to quality monitoring segment for guidance re how to sell products/services that improve customer satisfaction. Call centers need to be educated that speech is about improving customer satisfaction.

The Persona looniness continues: The Verizon 411 system continues to get bashed (see Verizon voice 411 system struggling in this issue), mostly because of the persona. Verizon is obtaining a conversion % approaching 30% which is technically quite good. Yet, the use of a persona is unnecessarily making this into a problem. This is simply the backlash from trying to be cute and not really understanding the needs of the caller. Callers to directory assistance are looking to quickly obtain the phone number of a person or business that they want to call. They aren’t looking to be entertained. When it doesn’t work perfectly, it offends them. Callers hate to repeat things and they hate it even more when they have to do it with a voice that is trying real hard to be friendly. This is phony nonsense and many callers are offended by it. Basic human nature. What are these people thinking! The VUI problems that Tellme/Verizon are experiencing, confirm that the implementation process that Tellme is using is fundamentally flawed. Discovering at this stage that the voice was speaking too slowly and saying a mournful "sorry" was totally inappropriate, simply confirms that the VUI design approach is wrong. The Verizon DA live agents should have designed the script. They know what to say and just how to say it including how fast it should be said. The fact that the Vernon DA live agents don’t like the automated system and claim that they do it much better is telling. The Vernon PR folks blaming the caller (customer) is cute and exactly what I would expect them to do. They would be better off in listening to the complaints of their customers and then focusing on meeting the needs of their customers (callers). The Telme folks identifying that the population of callers includes a tiny % that are weird enough to want to know more about the personality of the Persona is strange. Might be more interesting to speculate on whether these are the same folks that are pedophiles or sexual deviants. Meanwhile, some folks in the UK have "discovered" personas (see Persona Survey. Who Are You Talking To ? in this issue). Here I thought that the rest of the world had more sense and that this was just American nuttiness. Don’t know what is motivating FluencyVoice to hype personas, but I would suspect that it has something to do corporate greed and convincing naïve customers to spend a lot of money on this nonsense. Fluency has amassed hundreds of voices and can provide a customer with whatever type of persona they want. Just need to specify the voice: Chatty-sounding; professional-sounding; friendly-sounding; cold-sounding; bored-sounding; bossy-sounding; casual-sounding; arrogant; commanding; patronizing; shy; fawning; pleading; mean; serious; competent; or silly. Just imagine, customers can now waste time and money debating whether they should sound competent or friendly. You can get everyone in the enterprise, from the CEO to the janitor to assist in the VUI design process. If they don’t get it right the first time, they just keep doing it over-and-over. This would be funny if wasn’t so incredibly stupid. The only characteristic of the voice that is important is whether the caller likes the voice. Focus on this and you have a VUI that works.

When Outsourcing to Indians Makes Sense By Paul Stockford, President, Saddletree Research, Paul@Saddletreeresearch.com (480) 922-5949. At the end of last month, Dell Computer announced that it had stopped using a technical support center in India to handle calls from its corporate customers. After what is described by various sources as "an onslaught of complaints," Dell moved technical support for corporate customers to call centers in Texas, Idaho and Tennessee. Although Dell would not discuss the nature of the problems with the call center in Bangalore, India, media sources cite customer complaints regarding communications difficulties due to agents’ thick accents and their reliance upon scripted responses.

Earlier this year Saddletree reported the results of independent research that indicated that Indian call centers are now starting to display many of the same characteristics as their U.S. counterparts, including agent turnover as high as 40 percent, increasing absenteeism and a desire to job hop. Indian agents are showing signs of job burnout and an inability to maintain a work-life balance as they must work through the night in order to serve North American customers, must work on Indian holidays and take American holidays off and miss important family interaction as they leave for work as other family members are returning home.

As work-related health and cultural issues continue to mount in Indian contact centers, we believe the advantages of outsourcing to India will gradually disappear. That doesn’t mean, however, that outsourcing to Indians doesn’t make sense.

The Navajo Nation covers 27,000 square miles in northern Arizona, New Mexico and Utah. There are approximately 180,000 Native American Indians living on the Navajo reservation with a median population age of 22.5 years. According to a study undertaken by the Navajo government Division of Economic Development in 2001, 56 percent of Navajo people live below the poverty level and unemployment is at about 43 percent.

The capital of the Navajo Nation is Window Rock, AZ, located approximately 250 miles northeast of Phoenix, AZ and 173 miles west of Albuquerque, NM. The nearest city is Gallup, NM, located approximately 25 miles south of Window Rock. With a population of over 3,000 people, Window Rock resembles many other towns with retail outlets, lodging and restaurants, convenience stores, gas stations and other small businesses. 82 percent of Window Rock’s residents have completed high school and 25 percent have a college degree or higher. The mean age of Window Rock’s residents is 27 years old. Unemployment in Window Rock is nearly 14 percent, below the average of the reservation but nearly triple the U.S. national average.

For many Navajos, the idea of leaving the familiarity of the reservation and the unique aspects of their culture and cultural life is not an attractive option. Much of the population pursues traditional vocations such as farming and ranching and other occupations that allow them to stay on the reservation. The lack of infrastructure means that many residents still don’t enjoy modern conveniences that others take for granted. According to the "Census 2000" report from the Division of Economic Development, Navajo Nation, approximately 32 percent of homes on the Navajo reservation lack complete plumbing and 60 percent lack telephone service. It should be noted, however, that many Navajo families shun these conveniences in favor of living a more traditional lifestyle.

On the other hand, many leaders of the Navajo Nation are actively pursuing businesses with the objective of convincing them to explore the possibility of locating facilities on the Navajo Nation. Although there is a small percentage of the unemployed population that have stopped looking for work, similar to the U.S. population of unemployed as a whole, the majority of unemployed residents of the Navajo Nation welcome the opportunity of gainful employment. It is important to note, however, that many of the unemployed residents would rather remain unemployed than leave the reservation and break their extremely strong cultural ties to the Navajo way of life.

One of the issues that has faced businesses considering locating facilities within the borders of the Navajo Nation in the past is the fact that the Navajo Nation is a sovereign nation and is recognized as a domestic dependent nation under the protection of the United States. This status gives the Navajo Nation the right to make its own decisions and enforce specific laws that may or may not be in common with those of the United States. As a result, businesses of all sizes have been reluctant to locate on Navajo land for fear of losing control of some aspect of the business and having no legal recourse.

Today Navajo leaders are working to negotiate deals that have the dual objectives of benefiting the Navajo Nation while protecting the interests of the company locating on Navajo land. Although negotiating with the central Navajo government can sometimes be a lengthy process, the end result is typically a limited waiver of immunity that provides the business owner with full legal recourse in the event of a problem or dispute in the future.

In the case of Kayenta, AZ, a town located on the Navajo reservation about 140 miles north of Flagstaff, AZ, Navajo government officials have undertaken the task of establishing Kayenta as a municipality fully capable of making all decisions at the local level rather than at the central Navajo government level. Similar to other municipalities in the United States, Kayenta is able to negotiate it’s own business deals and offer protection to corporations and business owners that is identical to the legal protection businesses enjoy elsewhere in the United States. According to Kayenta Town Manager Gary Nelson, Kayenta’s local government status means that the town is empowered to do its own economic development and it is doing so by creating leases and other business agreements that are highly attractive to businesses willing to locate facilities within the township of Kayenta.

Similar to Window Rock, Kayenta has a relatively young median resident age, in this case 21.9 years. Kayenta has a population of 4,900 people, 73 percent of whom have completed high school and 14 percent of whom have completed college or higher. Unemployment in Kayenta is a staggering 20 percent.

Saddletree Research believes that the Navajo Nation, and particularly the towns of Window Rock and Kayenta, offers excellent potential for call center sites. Although communications infrastructure may be somewhat lacking in some areas, it is no worse than India was a few short years ago. Basic communications infrastructure is in place and working with the Navajo government to improve infrastructure would, we believe, be significantly easier than working with foreign governments and communications authorities.

We believe it would be relatively easy to recruit an agent workforce on the reservation that would meet basic education requirements, be willing to learn and eager to work. Most importantly, the people working in the call centers would be Americans. They would speak English, understand American culture, would not object to working an American work day or having American holidays off, have an understanding of the American consumer psyche and would be contributing to the American economic base.

Equally important is the fact that the Native Americans employed at call centers on the reservation would have the opportunity to contribute economically to their own community while growing both personally and professionally in a job that would not force them to leave the cultural comfort of a familiar way of life. We believe that the potentially high morale generated by this kind of employment opportunity would lead to a workforce with a much lower propensity to leave the job and a much higher potential to view the contact center as a lifelong career.

Businesses that locate call centers on the reservation would, we believe, enjoy a substantial pool of prospective employees from which to recruit, a high degree of loyalty among those hired and the satisfaction of hiring American workers from a population that has been largely ignored by American society. Although we have not done any research specific to tax benefits, it is likely that any business locating a call center on the Navajo reservation would also enjoy certain economic benefits sponsored by the U.S. government.

Saddletree Research believes that the bloom is off the offshore outsourcing rose. Eventually, outsourcing contact center jobs to India will create more problems that it solves. Now is the time for smart, and responsible, contact center and customer care executives to start looking inward once again for resources within the borders of the U.S. that will fulfill basic customer care requirements, contribute to the American economy and be cost effective in the long term.

We believe the creation of contact centers on the Navajo reservation is just the tip of the iceberg in terms of keeping Americans employed while effectively serving a North American customer base. There are Indian reservations throughout the United States and Canada that are in a similar economic situation as that of the Navajo reservation of northern Arizona. Saddletree Research believes that locating businesses on those reservations and putting the population to work will make more sense in the long run in terms of employee loyalty, job satisfaction, longevity and customer-care-with-a-smile than can ever be saved in the short term by paying agents in India $400 per month.

Before more companies have to do what Dell Computer did and pull jobs out of India due to dissatisfied customers, before more customer relationships are put at risk due to unintelligibly thick accents that frustrate callers, and before more American jobs are sent overseas, contact center operational alternatives should be explored. Sending contact center jobs to India is a short-term solution that is being employed by shortsighted managers. Saddletree Research believes that putting Native Americans to work in contact centers on the reservations offers a long-term solution that is not only economically viable and potentially profitable, it is the right thing to do.

adidas launches interactive point of sale experience to promote David Beckham Predator Pulse Collection To promote the launch of the new David Beckham Predator Pulse collection, adidas (Portland, Oregon) has partnered with Angel.com to create a unique shopping experience for the soccer fan. Starting in December, displays will be set up in approximately 100 Sports Authority and 50 Dick's Sporting Goods stores nationwide where the David Beckham collection will be merchandised. Each store account will have a custom toll-free number that consumers can call to hear from David Beckham himself about the features of his new collection. David Beckham informs the caller that his new range of Predator footwear and apparel is the ultimate combination of performance technology and style, allowing him to perform his best in every match. This display will take point of purchase to the next level, allowing adidas customers to hear what a superstar athlete thinks of a new product line. At the same time, it will provide valuable information regarding customer interest in the Predator collection, by tracking the number of people who call the toll free numbers from the stores. www.press.adidas.com Contact adidas America soccer PR: Evan Wiener 971-234-4647 Evan.Wiener@adidasus.com Mike Sweeney 703-770-1352 sweeney@angel.com

Voice Automation Helped Votes Count in ‘Decision 2004’ Election: Intervoice, Inc. (Nasdaq: INTV), and alliance partner Contact Solutions (Herndon, VA) announced their participation in NBC News’ comprehensive political coverage for the "Decision 2004" election. Given the controversies in 2000 that surrounded vote counts in several states and the dramatic increases in recent voter registration, NBC News exclusively broadcasted analysis of voter procedures and facilities using a Voter Alert Line application (1-866-MYVOTE1) that allowed voters throughout the country to immediately notify local polling officials about voting problems or irregularities at their local polling locations. This nationwide toll free voter alert line application was powered by the Intervoice Omvia platform and by the highly available, shared-service environment from Contact Solutions. If callers cannot successfully vote or don’t feel as if their vote was properly captured for any reason, voters can call this toll free application and have the opportunity to leave up to a 60-second message (in either English or Spanish) as to what issue was encountered. The message left by callers was captured as a voice file and delivered to NBC News along with statistics about the origin and volume of the calls via a web portal developed by Contact Solutions. At the end of the call, an option was also available to allow callers to transfer to the local election offices for information about their local polling location. A team of analysts examined the calling patterns and reviewed the content of individual calls for insight into local, regional or nationwide voting problems. On Election Day, November 2, 2004, MSNBC.com showed data collected through the Voter Alert Line, displaying real-time "hot spots" with a high volume of calls reporting voting problems. As a key component of the application, the management and monitoring tool from Intervoice, Application View, provided statistics in real time and users were able to view information on national, state and congressional district levels. Application View includes reporting processes and sophisticated tracking of user goals, events, number of callers, call duration, and other factors. For example, if calls in Florida started spiking, indicating voters were having widespread issues, analysts were able to access the web portal developed by Contact Solutions that created a map of problem areas as detected by the Application View tool from Intervoice. A state, region or city will light up and then the administrator can click on the blinking light within the portal to drill down to which county that is actually having issues. The administrator for NBC can then draw real-time statistics and retrieve .wav files to analyze the real-time voter activity in that county. The Voter Alert Line is funded by a range of non-partisan organizations and academic institutions including: The Common Cause Education Fund, The University of Pennsylvania Fels' School of Government, The Reform Institute, and The Hispanic Voter Project at Johns Hopkins University. The Making Your Vote Count site on MSNBC.com ( http://vote.msnbc.com/) will highlight reporting from all the NBC News programs, as well as provide voter registration information and sign up, voter technology interactive applications and the latest headlines on voting issues. www.contactsolutions.com Contact: Paul Logan, 866-979-3339

1, 2, 3...Sold! CreaLog Voice Portal Powers Germany's First Interactive Television Auction Channel: NMS Communications (NASDAQ: NMSS) and CreaLog announced that 1-2-3.TV ( www.1-2-3.tv ), the completely new German interactive television shopping experience, is using CreaLog's voice portal and CRM solution platform with NMS' Open Access platform ( www.nmscommunications.com/OpenAccess ) as the underlying architecture. For the first time, German consumers can determine their own prices for a variety of products by using 1-2-3.TV, an auction-like service described by CreaLog as similar to "eBay powershopping on TV." Like typical home shopping networks, viewers are presented with a selection of items for sale, such as clothing, sporting goods, cosmetics, home improvement items, and gifts, but because the price shown is only a minimum to start the bidding process, the similarity ends there. Customers can then offer their maximum price by telephone using the CreaLog voice portal and CRM solution, built on NMS' Open Access platform. CreaLog not only provides the voice portal but also the web and application servers used for data warehousing, commodity management, and fulfillment, as well as the planning tools used by 1-2-3.TV to manage and plan the product sales. In addition, all integration with the call center (Quelle.Contact) and the logistics company (DHL) is provided by CreaLog. Customers who register to bid with a 1-2-3.TV call center agent are entered into the database via the CreaLog Web interface (Web registration is also available). This process also automatically checks the customer's credit history so that delivery options can be adapted accordingly. CreaLog's voice portals can handle up to 100,000 calls per minute, a necessity for the massive call volume generated by the televised auctions, for which more than 20,000 people became registered bidders in the first month alone. 1.2.3.TV's business plan anticipates this figure will grow to 500,000 bidders in 2006. The CreaLog platform combines IVR, CTI, speech recognition in more than 40 languages and text-to-speech (TTS) in more than 20 languages with ACD, fax and E-mail in a single voice portal. It's already used in 13 countries, in more than 250 client installations. www.1-2-3.tv/index.php?id=37 for a live feed of 1-2-3.TV, which will be available until 12/15/2004. CreaLog is a provider of voice portal solutions in Europe, with references in 13 countries throughout Europe. CreaLog's VoiceXML-based and web-controlled voice portals are running at more than 250 leading customers in all industries. www.crealog.com . NMS Communications, Pam Kukla, 508-271-1611 Pam_Kukla@nmss.com colin_doherty@nmss.com Michael Kloos, CEO of CreaLog Dr. Andreas Buechelhofer, the second Managing Partner

Eckoh Technologies and Twenty First Century Communications win at NIE: Eckoh Technologies plc ("Eckoh") and Twenty First Century Communications, Inc. (TFCC) announced the launch of TFCC’s High Volume Call Answering (HVCA) service for Northern Ireland Electricity (NIE). This is the start of a three-year contract to provide these services to NIE. NIE is responsible for the regulated procurement, transmission, distribution and supply of electricity to 750,000 homes and businesses in Northern Ireland. It is the largest company within the Viridian Group plc. TFCC currently serve over 70 US electric utility clients covering over 69 million customer meters in the North American marketplace. TFCC’s HVCA service is the industry-leading power cut call handling solution. The HVCA® service has been developed and refined over 15 years and has been customised and re-deployed for the UK market supported by Eckoh who will exclusively host and manage the call handling platform on behalf of TFCC. The HVCA service will provide NIE with the ability to automatically process large volumes of calls resulting from power cuts. The service allows customers to inform NIE that they are experiencing a power cut, enables the capability of providing specific messages regarding the effected area, estimated restoration times by area, and includes call back options to confirm power restoration, as well as numerous other proven features. The service will utilize speech recognition technology to identify the customers by their postcode, and is being rolled out in phases across NIE’s service territory beginning in December 2004. Contact: Eckoh Technologies plc, Martin Turner, Chief Executive, Nik Philpot, Chief Operating Officer, Jim Hennigan, Managing Director, Speech Solutions Tel: 08701 100 700, Michael Skelton, Emergency Response Manager at Northern Ireland Electricity

Eckoh Awarded 3 Year Age Verification Contract for O2: Eckoh Technologies plc ("Eckoh"), has been awarded a 3 year contract to provide a user interface for age verification and parental control services from O2, a leading provider of mobile communications services. O2 are introducing age verification to ensure that commercial content services classified as 18 are only available to users who are old enough. Users will only be able to access these services if they have been successfully age verified. To verify their age, users will be able to dial an advanced speech recognition (ASR) service developed by Eckoh and hosted on their market leading call processing platform. The system will initially identify the caller by their mobile number and will use ASR to collect the details required from the caller to perform age verification to O2’s robust standard. On successful verification, the service will update the profile held by O2 to enable them to access content and services classified as 18. The system will also provide a simple way for users to configure O2’s new parental control service. Parental control is designed to help parents ensure the safety of their children by restricting access to Internet sites available through mobiles. This will be the first part of the solution provided by Eckoh to be delivered and will launch in December 2004. Contact: Eckoh Technologies plc, Martin Turner, Chief Executive, Nik Philpot, Chief Operating Officer, Jim Hennigan, Managing Director, Speech Solutions, Tel: 08701 100 700 Jonny Shipp, Head of Content Standards at O2

MM Group Handles Customer Request Peaks with Speech: Virtual agents are proving helpful in supporting MM Group staff handle peaks in customers requesting information, as callers welcome the new interactive voices over touchtone systems. These virtual agents, real voices operated by advanced software, have been used on a number of clients including: a government agency; Bradford Exchange; and Glade Air Fresheners. The services, provided by Fluency Voice Technology, include functionality to recognize callers’ name and address details, telephone and media codes, as well as, routing calls to appropriately skilled staff. For the government service, call completion rates have increased by over 30 percent, while the other services have collected over 50 percent more customer details than would have been achieved using other automated call handling systems. Rachel Robinson, business development director, MM Group,

Metaphor Announces Three New Customers for its Packaged Speech IVR Applications: Metaphor Solutions announced that three new customers signed up for deploying its packaged applications for their speech IVR requirements. The customers selected different applications on three different platforms. The first customer is Pull-A-Part, a chain of used auto parts and automobile junk yards in south eastern United States. Pull-A-Part deployed a complete suite of Retail Package Applications from Metaphor including directions, company information, pricing, inventory, consumer voicemails and phone and fax outbound notifications on the Microsoft Speech Server 2004 platform. Metaphor worked jointly with Definition 6, a system integrator and Microsoft Gold Certified Partner, on this implementation. The second customer, Wolpoff & Abramson, a law firm focused on collections, is deploying an outbound application suite on the Envox VoiceXML platform. Apptix, the third customer, and one of the largest Microsoft Exchange hosting providers, will be deploying Metaphor's SpeechOutlook application for providing voice access to Microsoft Outlook by phone to its corporate customers. SpeechOutlook will be implemented as a hosted solution on several VoiceXML gateways. Users will be able to use SpeechOutlook to name dial, read emails, send voicemails to emails, manage their contacts and manage their calendars, all from their existing Microsoft Outlook. And because SpeechOutlook works directly from Microsoft Outlook, users never need to transfer their current address book. www.metaphorsol.com michael@metaphorsol.com

Jet Multimedia buys 4000 telisma ASR ports and becomes Europe’s #1 hosting services provider for speech-enabled services: Telisma, announced that Jet Multimedia, leading hosting services provider in France for Internet, Wap, Audiotex, Minitel and telematics purchased 4000 Automated Speech Recognition ports. The Jet Multimedia investment is a result of demand from big businesses, including several companies seeking to rapidly enhance their telephony and voice services with speech recognition functionalities. The Jet Multimedia-telisma multimedia IVR platform already supports very large-scale applications for the general public: the first solution, already deployed for 18 months, includes no fewer than 1,000 simultaneous ports using telisma technology. Telisma’s speech recognition technology has demonstrated its performance with the interactive voice portal of TF1, the leading French television channel. The first thousand telisma ASR ports were deployed in phase one of the project, allowing callers to just say the title of their favorite service to be connected immediately. The portal was a real success. So in phase 2 of the project the decision was taken to deploy another ASR 4000 ports to provide an industrial answer to managing call peaks that can reach 1.5 million in two hours during popular, prime-time programs. For Jet Multimedia, this deployment will be the occasion to migrate to teliSpeech, telisma’s new generation of speech recognition technology, that opens up new opportunities for evolution of the service, namely in terms of free expression. Bruno Cunha, Marketing Manager at Jet Multimedia Hosting. www.telisma.com Contact: Pascale Garreau pgarreau@telisma.com +33(0_1 42 45 45 91

New health service contact centre, help line benefits British Columbians: TELUS (Vancouver, BC) has signed a 10-year, $20 million agreement to sub-contract services including contact centre technology and a speech recognition system to MAXIMUS BC HEALTH BENEFITS OPERATIONS INC., a MAXIMUS Canada company, to support that firm's delivery of program management and information technology services to B.C.'s health benefits operations. As part of the agreement with MAXIMUS BC, TELUS will provide: - a speech recognition system for a new 1-800 help line - managed contact centre technology and infrastructure - long-distance and 1-800 toll-free service - data server hosting. TELUS' speech recognition system uses caller information to determine the appropriate area to direct calls. Integrated Contact Centre Solutions will allow staff to answer these calls - as well as requests received by fax, email, web-chat, voice-over-web, voicemail and web callback - more quickly and efficiently. The B.C. Government is modernizing and improving administration of the Medical Service Plan and PharmaCare Plans. After a year-long procurement process, MAXIMUS BC was recently selected to provide program management and information technology services to the government. B.C.'s health benefits operations services include responding to public inquiries, registering clients, and processing medical and pharmaceutical claims from health professionals. TELUS has partnered with the B.C. Government on several technology-based service delivery projects. MAXIMUS CANADA INC. is the wholly owned Canadian subsidiary of MAXIMUS Inc. (NYSE:MMS). MAXIMUS INC. is one of North America's leading government services companies devoted to providing program management, consulting and information technology services. www.telus.com Contact: Shawn Hall, TELUS media and public relations, (604) 697-8176, shawn.hall@telus.com

GEOCOMtms and Vocantas Automate Home Delivery: GEOCOMtms (Atlanta, Georgia), a global provider of Fleet Optimization and Territory Design solutions, has teamed up with Vocantas Inc.( Ottawa, Ontario), a leading solution provider of computer telephony, IVR, and speech integration, to bring automated customer self-service and home delivery scheduling to the retail industry. The combined solution will enable customers to schedule delivery times automatically via the telephone and help retailers make more accurate delivery commitments. "Large retailers often have difficulty managing the logistics of home delivery," said Michael Nark, President and CEO of GEOCOMtms, "frequently using paper-based planning forms or simple spreadsheets to determine the order in which deliveries will be made each day. While this approach may work for smaller organizations," Nark continued, "at larger chains making hundreds of deliveries each day, the scheduling process quickly becomes difficult. The A.maze solution suite from GEOCOMtms enables retailers to more efficiently sequence, schedule, and route home deliveries, which not only reduces delivery related costs, but also improves customer satisfaction, because customers can be given more specific delivery time windows, rather than waiting at home all day for the delivery truck to arrive." Fully integrated into to the A.MAZE Suite, Delivery OnCall allows mass merchandisers to enhance customer service offerings through voice automation of customer-facing processes. Delivery OnCall is an Interactive Voice Response (IVR) system that enables customers to call in, and using their touchtone keypad, check when their purchases are scheduled to arrive in the retailers’ warehouse, schedule delivery to the home, and even request AM or PM delivery time. The system also performs outbound calls and automatically notifies customers that their purchases are ready for delivery. The partnership between GEOCOMtms and Vocantas leverages each company’s core strengths to create a solution that addresses both the unique challenges associated with home delivery and the industry’s focus on building strong customer relationships. For retailers, this means one more opportunity to enhance the customer service experience. GEOCOMtms has been helping shippers, couriers and carriers become market leaders by optimizing their pickup and delivery operations since 1999. www.GEOCOMtms.com Contact: Vocantas Inc., Randy Brown, (613) 271-8853 x596 randy.brown@vocantas.com www.vocantas.com GEOCOMtms D.Scott Beaver, Director, Product Marketing 770.803.0295x104 sbeaver@geocomtms.com

Viecore, Inc. Builds Automated Bus Information System for Palm Beach County Board of Education Viecore, Inc.(Upper Saddle River, NJ) has deployed a new solution for the Palm Beach County Board of Education. Viecore implemented an automated IVR solution to provide parents of students with special needs the precise time and location of student school buses. In most cases of Palm Beach County students with special needs, their parents are required to accompany them to and from their assigned bus route’s pick-up and drop-off locations. Buses are often delayed due to circumstances out of the bus driver’s control, such as a child requiring immediate attention, alternative bus routes, and fluctuating traffic conditions. Since the bus schedule is dependant upon so many external factors, Palm Beach County needed a bus location information system to provide parents with up-to-the-minute bus information. Viecore created the automated bus information solution with inbound and outbound dialing applications, providing exact location and estimated time of arrival (ETA) information on student buses. The new system will be built on the Avaya IR 1.2 platform, using a Viecore connector to integrate the Avaya platform with the bus’s positioning technology, the Geospatial Transportation system. Bus location information will be retrieved from the Geospatial system and communicated through the Avaya IR. The system will proactively provide bus information to parents through an automated outbound dialing application, or to a requesting parent through the inbound access application. Viecore is a leading systems integration firm specializing in enterprise level self-service solutions for customer contact centers. Founded in 1989, Viecore has developed deep vertical industry expertise, a large library of software assets, proven integration services, and premier partnerships with technology leaders in hardware and software. www.viecore.com 201-818-6600.Keri Chisholm, keri.chisholm@viecore.com

Viecore, Inc. to Create Speech-Enabled Eligibility and Benefits Self-Service Solution for Medical Mutual of Ohio Viecore, Inc. .(Upper Saddle River, NJ) announced that they have been selected to build a new solution for Medical Mutual of Ohio. Viecore will develop an enhanced speech-enabled eligibility and benefits solution for medical, dental, and vision Providers to streamline call center operations and improve customer service levels. Medical Mutual of Ohio selected Viecore to conduct a consulting-based assessment of their call center operations. The Viecore Application Analysis Process (VAAP) was utilized to determine the best case for enhancing existing DTMF applications with speech-enabled customer self-service automation. Based on the results of the VAAP study, Medical Mutual of Ohio contracted Viecore to design an eligibility and benefits solution that utilizes speech-automated self-service technology to boost call center efficiencies and streamline operational process for improved satisfaction. Viecore will create the enhanced Medical Mutual of Ohio solution for medical, dental, and vision Providers using directed dialogue speech technology. The Viecore solution will enable Providers to speak requests for eligibility and benefit information on members. Additionally, the solution will provide the mailing address and contact information for pre-certifications. Viecore will integrate the solution with Medical Mutual of Ohio’s CTI platform for effective call routing with successful screen-pop delivery. www.viecore.com 201-818-6600.Keri Chisholm, keri.chisholm@viecore.com

Viecore, Inc. Contracted by Sedgwick to Automate Claims Processing with Speech Recognition Technology: Viecore, Inc. .(Upper Saddle River, NJ) announced that they have been selected to build a new solution for Sedgwick Claims Management Services (CMS) a leading insurance claims processor for major corporations. Viecore will work with Avaya and ScanSoft to implement a speech-automated IVR solution to provide Sedgwick CMS client, claimants, and providers with convenient access to information and transaction processing 24x7. As a third party administrator, Sedgwick CMS handles a massive volume of contact center interactions annually servicing many customers, each with their own specific business rules and requirements. With the recent addition of a major account to the customer base, Sedgwick CMS has an urgent need to address additional customer needs without sacrificing customer service levels. Sedgwick CMS selected Viecore, Avaya, and ScanSoft to map out a solution that meets their unique requirements. Viecore’s Consulting Services Organization employed the Viecore Application Analysis Process (VAAP) to determine how and where Sedgwick CMS should invest in speech to meet its business goals and objectives. Together, Viecore, Avaya, and ScanSoft will build a speech-enabled self-service claims management solution for Sedgwick CMS to provide clients, claimants, and providers with faster, easier, more convenient access to information and transaction processing 24x7. Using Avaya’s Interaction Center (IC), the claims management system will collect data from the caller and simultaneously transfer the call and information via screen-pop to an agent for handling, providing a robust reporting structure around the claims processing calls. By minimizing touch-tone menus and offloading routine calls from busy agents, the solution will streamline the call flow process, allowing agents to focus on more challenging or complex transactions for lowered cost of ownership, improved operating efficiencies, and increased customer satisfaction. www.viecore.com 201-818-6600.Keri Chisholm, keri.chisholm@viecore.com

Aer Lingus Automates Flight Reservation Calls Using Voxify Automated Agents: Aer Lingus Airlines, Ireland's national carrier, and Voxify (Alameda, CA) announced a deal whereby Voxify has been charged to support the airline's phone reservations. This includes fare shopping, bookings, itinerary changes, seat assignments, and reconfirmations. Aer Lingus turned to Voxify because the company's technology models the conversational intelligence of human agents, ensuring that its call center customers will continue to receive excellent service while the airline transforms itself into the first transatlantic, low fare carrier. "We are committed to giving our customers the best fares across the Atlantic. To accomplish this, we have to find unique ways to lower our own cost structure while maintaining customer service," said Jack Foley, Aer Lingus' Executive Vice President, North America. "With Voxify, we will be the first low cost transatlantic carrier to offer end-to-end reservations through an Automated Agent. When we began this process we had strict requirements to create a high-touch experience even as we pursued becoming a low-cost carrier. Voxify's technology, with its conversational intelligence, provides us with the best technological solution to handle important transactions, including reservations." As Ireland's national airline, Aer Lingus North America handles more than a million calls a year within the United States. Internal studies demonstrate significant improvements in overall agent availability by providing Voxify Automated Agents. "In handling these calls Voxify's Automated Agents, are allowing Aer Lingus travelers to book their reservations 24 hours a day, seven days a week, without ever waiting on hold. Moreover, because the Automated Agents are hosted and maintained by Voxify, we are able to significantly reduce the per-call cost of these calls," said William Kane, Vice President, Information Technology at Aer Lingus Airlines. Automated Agents are one of several cost-cutting initiatives Aer Lingus has undertaken to position itself as a pioneer in creating low fares on transatlantic routes. So far the effort has paid off, with significant savings being passed on to customers. For example, the airline recently revamped its entire transatlantic business and leisure fare structure. Premier business class fares have been reduced by over 50 percent and lead-in economy fares have been reduced by 20 to 30 percent. When all Automated Agents are up and running the Agents will provide a wide number of self-service choices for customers calling in. And for Aer Lingus management, the Automated Agents will provide detailed reports on traveler habits and preferences, delivering important information for training and marketing. 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 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. www.voxify.com Contacts: Hollis Chin, Voxify, 510.545.5015 hchin@voxify.com

PostFinance chooses Swisscom Systems for telephone self-service banking solution on Voxpilot VoiceXML platform: The days of cumbersome touch-tone menus for self-service solutions are over – at least for customers of the Swiss financial institution PostFinance. From June 2005, they will no longer use the buttons on their phone to get account and finance information – instead they will simply use their voice to inquire about their account balance, fund prices, or to report a stolen bank card. The application, deployed on Voxpilot’s carrier-grade Open Media Platform, will be available in four languages (French, English, German and Italian) and will be protected by a user number and PIN code. With this telephone self-service, PostFinance offers its customers a much easier and quicker way to access account information. The service promises to increase customer satisfaction due to its user-friendly interface and 24/7 access to customer-critical information. Swisscom Systems put together the project for PostFinance, the financial arm of the Swiss Post Office. Insonic AG, a Swiss specialist for speech solutions, will develop the speech application, which is expected to answer more than 2 million calls per year. 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. www.voxpilot.com . Swisscom Systems is a specialist in the area of inhouse communication solutions for business customers and is able to offer a broad portfolio of services, ranging from planning, consulting, implementation and maintenance to different financing options such as purchasing, leasing and outsourcing. Contacts: Voxpilot: Dr. Michaela Heigl, Michaela.Heigl@voxpilot.com +33 4 93 00 45 92. Swisscom Systems: Dieter Wahlen Dieter.Wahlen@swisscom.com tel: +41 1 294 81 80

Apptera Adds New Packaged Speech Apps For Financial Services: Apptera has added modules for account enrollment, account payment, and mortgage rate quotation to its suite of packaged speech applications for the financial services industry. Apptera has added three new modules to its Financial Services Suite of packaged speech applications. The new apps automate a range of customer service requests, including establishing new accounts, making credit card and loan payments, and receiving mortgage rate quotes: *Apptera Account Enrollment leads callers through the process of applying for a loan or enrolling in a new account—asking the same questions as live agents, but, the company says, completing the enrollment process in just half the time. The speech application includes series of configurable qualification questions that financial institutions can enable or disable instantly using Apptera's One Click Configuration system. Account Enrollment integrates with CTI (computer telephony integration) systems to hand-off captured information to a live agent at any point during the call as necessary; *Apptera Account Payment lets customers make recurring and one-time credit card and loan payments by phone. Apptera says that the application eliminates hold times and reduces agent costs by voice-automating the entire account payment process; *Apptera Mortgage Quote helps provides automated rate quotes around the clock without the need for mortgage specialists. The application provides quotes for popular mortgages and custom quotes based on answers gathered from a set of configurable qualification questions. The application also offers personalized attention, with refinance calculations detailing overall savings to callers. Apptera's Financial Services Suite, which the company introduced earlier this year, lets banks and credit unions of all sizes develop speech-enabled interactive voice response (IVR) systems for customer service and revenue generation. The VoiceXML-based applications are platform-independent and ship with a browser-based studio for customization. Other modules in the suite provide packaged speech applications for account information, funds transfer, the location of ATMs and branches, and name and address capture. Apptera's Financial Services Suite are available as both on-premise and hosted deployments on a variety of leading IVR and speech platforms. Apptera's packaged speech applications are offered directly by Apptera and also through the company's platform and hosting partners, including Convergys, Intervoice, ScanSoft and SandCherry. www.apptera.com Contact: Jodi Guilbault, (650) 635-0617 jguilbault@apptera.com

Edify gets VoiceXML certification: Edify is announcing that its Edify Voice Interaction Platform (EVIP) 9.0 system for deploying speech activation applications has received VoiceXML 2.0 Certification from the VoiceXML Forum. The certification ensures interoperability between applications conforming to VoiceXML 2.0. EVIP is for deploying speech-automated self-service systems, where the application answers the phone and uses speech recognition, according to Edify. Customers of the company can operate VoiceXML-compliant solutions and native Edify voice solutions on the same platform, the company said. VoiceXML serves as a markup language to enable developers, platform vendors, and tool providers to reuse code and more easily develop applications and tools that can interoperate with other VoiceXML-based systems, according to Edify. The specification is intended to eliminate the need for proprietary markups. www.edify.com Cynthia Dimidik, Edify Corporation: (408) 486-1791; cynthia.dimidik@edify.com

Persona Survey. Who Are You Talking To? A new service offering companies the chance to identify the voice of an agent that best suits their business is launched today by speech recognition specialist Fluency. The "Persona Survey", available via the firm’s website allows companies to get an insight into this service by matching their organisations profile to a sample of agent voices. What makes these agent voices different is the fact that they are not real people – marketing professionals craft them as ideal profiles of people who best suit companies such as banks and utilities and are part of the speech recognition revolution changing the way businesses interact with their customers. The survey is used by Fluency to demonstrate the complex process involved in selecting these virtual agents and includes virtual people such as: Patsy Taylor; Andrew Warner; Brian Jones; Caroline Wilson; and Jessica Long. These virtual agents, real voices operated by advanced software, are already out there talking to us everyday when we pay our electricity bills, enquire about bank balances, or even book a train or coach ticket. The whole area of voice branding has been a blind spot for many years – names, logos, and visual brand presentation are very well established but many companies have yet to consider the brand impact that the personality or "persona" of an automated voice system can have on their business. By answering a few brief questions the Persona Survey helps companies identify a suitable persona from a small sample of the hundreds available from Fluency’s catalogue. Virtual agents provide a vital point of contact between companies and their customers and as such have the potential to either support or undermine that company’s brand. The type of voice that is used with the system, and the tone it adopts when interacting with callers will have a fundamental impact on whether the system is ‘on brand’. Voice selection and voice branding are relatively new areas and organisations are only now realising their importance. Barnes added: The last part of the process involves making the conversations seem as real as possible by introducing Earcons or non-speech audio. These sounds communicate to the caller by giving them clear signal as to what is happening, such as a keyboard tapping sound when their information is being entered in a system. This also allows brand reinforcement through audio "logos", music and other relevant sounds. Successful speech applications allow callers to complete their transaction in the system, or can transfer callers to a live agent as appropriate. As such they form an integral part of the contact centre, delivering high customer satisfaction through the use of intuitive Voice User Interfaces (VUIs) guiding the caller through a natural conversation. The speech application’s persona provides a consistent experience with other customer touch-points, helping to deliver a cohesive brand. Contact: Nigel Jones, The Purple Tangerine +44 (0) 7980 213 122/+44 (0)1233 712 899 nigel@purpletangerine.com

Paraxip Technologies Unveils New IP Telephony Products: Paraxip Technologies (Montreal, Canada) announced the availability of release 1.1 of its Paraxip Gateway product and launched Paraxip IP Record, a SIP-based call recording solution. The Paraxip Gateway is a software solution based on open, off-the-shelf hardware components, specifically targeted at the Interactive Voice Response (IVR) and call center market. The release 1.1 of the product introduces multiple enhancements including: *Call Forking – Paraxip's product architecture allows associating multiple SIP call legs with every PSTN lines. This gateway feature enables scalable and reliable call recording and call monitoring in the IP Call Center. *Call control enhancements – Multiple enhancements to Paraxip's advanced call control feature set are introduced in this release such as: advanced call progress analysis improving the return-on-investment on outbound dialing applications, fully supervised call transfers on hook-flash, added signaling protocol support including E1 PRI, and more. *Improved manageability – This release introduces numerous manageability enhancements including an extended programmable web service interface and the addition of an embedded web server that enables remote monitoring through a browser-based graphical interface. The company also released its Paraxip IP Record product, designed to provide a cost effective and flexible call recording solution in the contact center. Contact info@paraxip.com +1 (514) 288.7111

Vocalabs New Book availabe: In January 2005, Vocalabs will introduce their new book: Gourmet Customer Service: A Scientific Approach to Improving the Caller Experience Quality service has a direct impact on the bottom line. Call center technology has to be deployed to provide a good experience rather than just saving money. Findings presented in the book include: *Good service isn't more expensive than bad service. *Customers often prefer automated self-service, if it meets their needs. *Call centers can be more important to brand image than advertising. There are many reasons testing is often inadequate in a call center, and this book is intended to help companies understand both the "why" and the "how" of measuring the quality of customer service. Gourmet Customer Service is about using data gathering, experimentation, and scientific methodology to provide your customers with the best and most cost-effective service possible. We focus on how, when and what types of testing are needed to achieve that goal. Techniques used in measuring and improving customer service quality are the ingredients, which get combined into recipes for different situations you may face. Preorders received before January 15, 2005 qualify for the special Preorder Price of $14.95 per copy. After January 15, the price will be $22.95. pleppik@vocalabs.com

Acapela speaking for MajorCall Eureka European project. EUREKA, a pan-European network for market-oriented, industrial R&D has selected Acapela to partner on MajorCall project focusing on Customer Relationship Management. The MajorCall project is dedicated to redesigning business processes to develop a new generation of CRM, integrating speech technology with communication, marketing and customer-related services. Acapela will provide its high-end speech solutions (ASR & TTS) and develop specific modules to make speech a key component of the new CRM generation with cutting edge pre processor module, emoticons and signature management, and automatic language recognition. Other modules such as diacritics restoration, correction of misspellings, correction of typing mismatches will also be at the heart of the project. Created as an intergovernmental Initiative in 1985, EUREKA aims to enhance European competitiveness through its support to businesses, research centres and universities who carry out pan-European projects to develop innovative products, processes and services. www.eureka.be www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471

Diaphonics Provides Biometric Voice Verification for National Association for Variable Annuities’ Electronic Signature Initiative: Diaphonics (Halifax, Canada), a provider of security solutions based on voice verification, has partnered with Signix (Atlanta, GA), a leading digital signature company, to provide a platform for the electronic signing of legal documents for the NAVA E-signature working group. This working group, which includes representatives from major insurance carriers, distributors and securities firms, recognizes that electronic signatures can reduce paperwork, leading to greater efficiencies and reduced costs, and has therefore established a simulation lab to test various electronic signature solutions. Diaphonics’ Spike Server adds voice verification to Signix’ hosted solution for Public Key Infrastructure (PKI), so that a user’s identity can be verified over the telephone as part of the electronic document signing process. Spike Server verifies a caller’s identity by matching a sample of the caller’s voice to a saved voice profile, or "voiceprint". According to NAVA, some of the companies that are participating in the simulation include AEGON Insurance Group, Allstate Financial, American Express Financial Group, Finetre Corporation, Genworth Financial, ING U.S. Financial Services, MetLife, Minnesota Life, Nationwide Financial and Symetra. Diaphonics is a provider of security solutions based on voice verification. Today, financial services companies and government agencies rely on Spike Server to verify the identity of callers with voice verification, record voice transactions and create a searchable audit trail. Signix is a digital signature solution provider. The Company's technology allows online visitors to a website to be rapidly authenticated and issued digital signing credentials, enabling them to apply legally-binding signatures to documents utilizing the browser from any PC. Signix offers an advanced implementation of PKI, or public key infrastructure, which makes this highly secure technology practical and cost-effective for a wide range of new applications. NAVA ( www.navanet.org ) is a not-for-profit trade association that provides a variety of services to the industry including educational forums, research, and conferences aimed at furthering the development and understanding of fixed and variable annuities, income annuities and variable life insurance. NAVA also maintains and supports an educational website for consumers at www.RetireOnYourTerms.com www.diaphonics.com www.signix.com Contact: Jeremy Bernard, Diaphonics, (902)446-3671, jeremyb@diaphonics.com

Telstra scores $185M Centrelink voice contract: After a six-month delay Centrelink has finally awarded its managed voice services contract to incumbent Telstra in a contract worth $185 million over five years. The deal covers Centrelink's voice services, call centres and a range of telecommunications-based services ranging from interactive voice response, voice recognition and SMS-based services. It also covers all voice telephony services at the Department of Family and Community Services (FACS) and CRS Australia at more than 1000 sites across the country. Centrelink holds the mantle of having Australia's single, largest call centre operation with 26 call centres across the country. The agency's call centre functions play a pivotal role in a variety of ad hoc government campaigns and crisis management capabilities ranging from educating social security recipients to dealing with crises such as the Bali bombing and natural disasters. The tender award also notably sells the new technologies such as voice recognition to enable customers to better navigate their way through the human services bureaucracy. One expected winner is voice recognition application vendor Scansoft, which has found a range of its applications deployed to allow students on government benefits to file automated income reports using voice. Scansoft regional director Peter Chidiac said the announcement "is definitely good news for us" adding that many other government agencies were looking at speech recognition. One area not covered in the announcement is any possible deployment of VoIP technology or where it would fit into Centrelink's current technology framework. A spokesman for human services minister Joe Hockey said while the announcement did not include VoIP, it may be looked at by Centrelink as the need arose.

TuVox Announces $15 Million Financing Led By Norwest Venture Partners:. TuVox (Los Altos, Calif) has closed a $15 million Series D equity financing in an over-subscribed round led by new investor Norwest Venture Partners (NVP). All of the company's previous investors, including Foundation Capital, Granite Ventures and Adobe Ventures participated in the round. TuVox will use the new capital to fuel aggressive expansion in sales and market awareness in North America and Europe, and will be opening its European Operations in the U.K. in 2005. In addition, TuVox will accelerate the delivery of additional pre-built speech applications. Strategic investment motivated by increased demand from mass media for speech recognition. Strategic investment motivated by increased demand from mass media for speech recognition. TuVox also announced an 800 percent increase in revenue during fiscal 2004. 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 or call 1-866-257-1830.

Unveil Technologies and Genesys Form Strategic Relationship to Deliver Conversational Speech Self-Service Systems: Unveil Technologies, Inc. (Waltham, MA) and Genesys Telecommunications Laboratories, Inc., an Alcatel company (NYSE: ALA, Paris: CGEP), announced that Unveil has joined the Genesys InterActs Partner Program as a Strategic Member. As part of the agreement, Unveil can resell both the Genesys 7 contact center software and the Genesys Voice Platform in conjunction with its Conversation Manager solution. Unveil’s Conversation Manager speech self service applications run on the Genesys Voice Platform, an open standards-based IVR solution that allows customers to leverage existing Web applications and infrastructure for voice self-service via any phone. When combined with the Genesys 7 suite, the solutions enable companies to seamlessly transition customers between self-service and live agent assistance, resulting in reduced costs as well as improved customer satisfaction. Unveil Conversation Manager is voice application software that enables enterprise call centers to rapidly build, deploy and maintain high performance speech applications, or "virtual agents," that increase automation rates, improve customer service and reduce the total cost of ownership. Unlike traditional speech self service applications, Unveil virtual agents: · Engage customers in unscripted, open conversations that enrich the customer experience and automate complex transactions; · Integrate with live agents to extend self-service transactions and improve caller satisfaction (Conversation Assist), and; · Utilize real customer interactions to dynamically improve over time, reducing the total cost of ownership for speech (Adaptive Learning). The Genesys InterActs Program is designed to provide business development, technical, educational and marketing tools and resources to participating system integrator, value-added reseller and managed service provider partners worldwide. Unveil provides enterprise voice application software that enables call centers to rapidly build, deploy and maintain high performance speech applications that reduce costs and improve customer satisfaction. www.unveil.com www.genesyslab.com Contact: Unveil Jim Williams, 781-250-1105 jwilliams@unveil.com Genesys, Dana Dye 650.466.1078 danadye@genesyslab.com

VOICE.TRUST couples telisma’s speech recognition with its speaker verification in order to better secure transactional calls: VOICE.TRUST and Telisma announce their technological and commercial partnership to address the European market. The joint solution will rely on VOICE.TRUST speaker verification technology and telisma’s speech recognition technology, teliSpeech. The secret behind the VOICE.TRUST-telisma joint solution is that it relies on both voice verification and speech recognition technologies to optimize security. This dual approach is in fact at the core of the companies’ technical partnership- VOICE.TRUST technology recognizes the caller via his voice print, while teliSpeech recognizes that the password or personal identification numbers he is uttering are the right ones. Speech recognition is further used in the dialog phase to get conversational vocal access to the telephone services secured by the system. The VOICE.TRUST.FIVE System featuring teliSpeech is available right now in all European languages. www.voicetrust.de www.telisma.com Contact: Pascale Garreau pgarreau@telisma.com +33(0_1 42 45 45 91 Andreas.Battenberg@Voicetrust.com Tel.: +49 (0)89 - 127 16 0

Witness Systems to Acquire Blue Pumpkin Software: Witness Systems (NASDAQ: WITS), announced that it has reached an agreement to purchase Blue Pumpkin Software, Inc., an industry leader in enterprise workforce management solutions, for approximately $75 million. Approximately $40 million of the purchase price is being paid in cash and the remainder in Witness Systems stock. The merger reinforces the company's position in the burgeoning global contact center workforce optimization market, and is expected to be accretive to earnings for the 2005 calendar year. Uniting Witness Systems' established lead in customer interaction recording with Blue Pumpkin's market strength in workforce management offers customers a compelling and seamless set of applications to help optimize workforce performance and capture customer intelligence. The combination will create a unique leadership position in the two anchor segments of the broader workforce optimization landscape (quality monitoring and workforce management). Blue Pumpkin has developed a strong reputation in the small- to medium-sized business (SMB) segment of the market, which is expected to be the fastest growing segment over the next five years. www.witness.com Contact: Anne Patton, Witness Systems 770.754.8656 apatton@witness.com

Avaya Ships Modular Messaging 2.0: messaging application. New features include a Web client, integration or IBM Lotus Domino, and better integration with Microsoft Exchange that includes greater scalability capabilities. Avaya Modular Messaging 2.0 can support up to 20,000 users at multiple locations without requiring separate servers at each site, which accounts for much of the cost savings users realize from the application. The system uses IP networking to transport messages between locations, integrating communications systems from different vendors via open standards-based technology. Another improvement in this version allows users to access their messages from their PCs as well as their phones via an Avaya software plug-in that connects to Outlook and Lotus Notes interfaces. The system also allows workers to use Web browsers to tap into the application, according to the company. Version 2.0 is speech-enabled. The Avaya Unified Communication Center allows mobile employees to access and manage messages, calendars, directories and tasks, as well as make phone calls and set up conferences, from any phone. Avaya Modular Messaging 2.0 can support up to 20,000 users per system. With IP telephony and IP networking, a single system can support multiple locations without requiring separate voice mail servers at each location, providing more application power for all users and delivering significant cost efficiencies. With Modular Messaging 2.0, end users can now access their messages from their personal computers (PCs) at home or at the office as well as from their phones. A simple software plug-in from Avaya enables users to access their voice messages through the popular Microsoft Outlook and IBM Lotus Notes interfaces, making it easy to manage voice mail, e-mail and fax communications from the user's laptop or desktop PC. The system also allows workers to use common Web interfaces, such as Microsoft Internet Explorer or Netscape browsers, to surf into Modular Messaging and listen, reply, forward and send voice, fax or text messages from any Internet-ready PC. Speech access software can make messaging even more easy and convenient to use. www.avaya.com

First Impressions Virtual Receptionist Ready for Calls: Invores Systems, Inc. (Long Island, NY) announced the launch of First Impressions, a new virtual receptionist software package for business. Callers are greeted by the virtual receptionist and asked who they are calling…just like a "real" receptionist would ask them, and then the caller is transferred to that person or department. The First Impressions virtual receptionist greeting can be customized through an intuitive browser interface giving companies the ability to tailor it to their exact needs and personality. Invores Systems plans to market its First Impressions virtual receptionist through a network of authorized reseller partners. About Invores Systems: Invores is a privately held Long Island, New York based software development company focused on Interactive Voice Web speech-enabled customer self-service solutions in commercial, service provider and government organizations. Invores Systems offers a comprehensive suite of licensed software and services, designed to automate customer contact business processes through the telephone while capitalizing on existing web infrastructure investments. www.invores.com Contact: Debra Altshuler, VP of Marketing, 516-868-5455

HeyAnita Launches Rapid Message Service v2.0: HeyAnita Inc. unveiled version 2.0 of its Rapid Message Service (RMS) software. RMS is text messaging without the text, enhancing traditional texting by allowing users to speak a message rather than typing it on their phone, regardless of the handset or network they are using. In addition to new international and enterprise features, RMS v2.0 further simplifies the manner in which users provision the service and manage their RMS contacts. The new operator configurable features of RMS v2.0 include Web-free management of the RMS Address Book, Web-free self-provisioning, enterprise administration tools and international language support. Web-free management of the RMS Address Book includes Receiver Auto-Population, which automatically adds contacts to the user’s RMS Address Book when they receive an RMS from another registered user who is not already listed as a contact, and Send N’Add, which allows users to add new contacts to their address book via SMS. Web-free self provisioning provides recipients of RMS messages with the ability to sign up for the service with the touch of a key on their mobile phone, without requiring them to visit an online Web site. RMS v2.0 supports five languages including: U.S. English, French, German, Spanish and U.K. English. The international configurability of v2.0 covers localized audio prompts, text notifications and Web sites as well as phone number management. Additional languages will be supported in the near future. In addition to other enhancements, RMS v2.0 also includes the ability to disable replies to RMS sent via distribution lists. This is helpful when sending to a large recipient group, such as a sales force, and you don’t want to receive responses to your message. The RMS solution, which works across mobile operator networks without requiring special handsets or handset software, works in the following way: -- The sender selects one or more recipients by saying a person's name or by keying in the recipient's mobile phone number. Then, the RMS message is recorded. -- A text notification, which includes the sender's name, is sent to the recipient. -- While viewing the text notification, the recipient presses a designated key to hear the RMS and can reply to the message with the touch of a button. -- Anyone can receive and reply to an RMS, even if they are not subscribers to the service, regardless of operator network or handset technology. RMS is currently in trials with major wireless operators in Europe and North America, as well as being trialed by several thousand individual users. www.heyanita.com Mark Willingham, vice president of marketing for HeyAnita

iVoice Speech Enables Bergen County Utilities Authority: iVoice, Inc. (BULLETIN BOARD: IVOC) , announced that the Bergen County Utilities Authority of Little Ferry, N.J. has installed an iVoice Speech-Enabled Auto Attendant with directory service. The county’s Utilities Authority implemented a solution that includes iVoice Speech-Enabled Auto Attendant installed in conjunction with the Avaya PBX phone system. By installing the iVoice Speech-Enabled Auto Attendant, the authority’s customers can call in and speak the name of the department they are looking for and be transferred immediately. The Auto Attendant was sold and installed by iVoice partner PhoneXtra. The Bergen County Utilities Authority (BCUA) is a public utility providing sewage disposal for 46 municipalities in Bergen County and solid waste services for 70 municipalities in Bergen County. The BCUA board is comprised of commissioners appointed by the county executive with the consent of the Board of Chosen Freeholders. PhoneXtra is an Avaya "platinum" certified business partner and maintains the distinction of having an Avaya Loyal "Diamond-Dealer" status. www.ivoice.com Frank Hawkins or Julie Marshall, Hawk Associates, at (305) 852-2383, e-mail: info@hawkassociates.com

IP Communications Lets Winthrop and Weinstine P.A. Provide a Consistent, Highly Available Communication Channel Between Attorneys and Clients at a Lower Cost: Spanlink Communication (Minneapolis, MN), a leading IP communications and contact center solution provider, announced a success story for a 250 phone IP Communications solution at Minneapolis-based law firm, Winthrop and Weinstine. Winthrop & Weinstine, P.A. is a full-service, entrepreneurial, commercial law firm dedicated to advancing the best interests of its clients and community through the timely delivery of legal services of exceptional quality, innovative character and unsurpassed value. When Winthrop and Weinstine consolidated their offices from two locations into one building in Minneapolis, they did a feature and cost analysis between taking their existing TDM-based switches and moving them into the new, single site or moving to a new IP-based phone system. The Winthrop IP Communications solution includes Cisco CallManager, Cisco Unity, a Spanlink Click-to-Dial Application integrated into Outlook, and wireless phone service for attorneys and staff. According to Craig Wilson, MIS Manager, "Based on the required feature list from end users, we could have purchased another TDM system or an IP system. They both provided the core features the attorneys wanted. But with the IP system, we have the ability to do more. The attorneys live on their office phone and cell phones, and anything we can do to increase their phone efficiency has a tremendous benefit." For example, Spanlink developed an application for one-click dialing from Outlook. According to Wilson, "The application lets you right click on the number within Outlook contacts and dial, which is a huge time-saving benefit for the attorneys. "Another time-saving application is unified messaging. With Cisco Unity, attorneys can work through voice and email messages very efficiently. They can even respond to an email with a voice mail. Today, the majority of our attorneys are listening to their email messages on the phone when they are out of the office." Wilson added, "For me, time was a big advantage of the IP solution, because the last thing you want to do is be down. Moving two systems into one, we would have been down for about four days. Purchasing a new IP-based phone system offered the benefit of no real downtime. We also have the advantages of less cabling, and the ability to support the system internally. With our previous solution, we were spending $45K a year to contract a vendor for adds, moves, changes. It cost us a minimum of $150 each time just to move some wires. Now, we are able to bring that all in house and support it with our existing IT staff." "The biggest issues of this big move had do with how the furniture was arranged - not the new phone system," said Wilson. "We had training two weeks prior to the move, which was a great advantage. We had IP phone stations set up around the building, which enabled people to go in and test calls to get them used to the new system before they were actually using it. The learning curve for the new system was very quick." The Results Said Wilson, "I need to provide my clients with the best possible service in order for them to efficiently conduct their job. The IP phone system has enabled me to do that. Today, we have an environment that is flexible enough to support high-end and low-end users - one that provides a consistent, secure, highly-availability communications vehicle between my internal customers, the attorneys, and our external clients." "With the IP system, we had to do less wiring, we were able to train prior to moving, and we have the core features we need. Now, we can take it to the next level to continue to increase efficiency by adding additional features when we need them. This type of advanced or additional options are simply not available with a TDM system." Spanlink Communications is a leading solution provider of IP communications and contact center solutions. Spanlink delivers next-generation customer interaction and IP Communications System Management products that maximize efficiency in a pure-IP or migratory environment. Spanlink is one of the leading reseller-integrators of Cisco Systems IP Communications and Contact Center products, including Cisco CallManager, and Cisco Intelligent Contact Management (ICM), Cisco IP Contact Center (IPCC) Enterprise, and Cisco Customer Voice Portal (formerly ISN) and IP-IVR. www.spanlink.com Contact: Kristen Jacobsen Spanlink Communications 763-971-2147 kristen.jacobsen@spanlink.com

Neoris Selects VoiceGenie for Dynamic Truancy Alert Management System in Educational Sector Performance Capabilities of VoiceGenie Speech Platform Unrivaled in Reliability Evaluation: VoiceGenie Technologies Inc. announced the selection of their VoiceXML 2.0-certified platform, NeXusPoint, for Neoris’ speech-driven truancy alert management application deployed at the Fort Worth Independent School District (FWISD) in Texas. Neoris, a leading provider of IT consulting services, partnered with FWISD, the 4^th largest school district in Texas with 143 schools and more than 80,000 enrolled students, to reduce the district’s truancy rate and implement a district-wide intelligent communication messaging strategy. In addition to preventing students from skipping school and potentially dropping out or contributing to urban crime, keeping students in the classroom represents a significant financial opportunity for the FWISD. The State of Texas appropriates funding to each school district based on the number of students in class each day during the regular school year; therefore, every truant student costs the FWISD US$ 15 per day. Neoris’ ActiveSuite Alert Manager solution, deployed on VoiceGenie’s NeXusPoint platform, notifies parents and guardians as well as law enforcement authorities when a student is truant from school and provides a confirmation when the intended recipient receives a message. It is estimated that the FWISD will reduce truancy district-wide by more than 10% over the next two years, resulting in additional funding from the State government of up to US$ 700,000 each year. www.voicegenie.com/ . Neoris is a U.S.-based IT Services company recently ranked as the 4^th largest IT consulting firm and 5^th largest custom application development company in Latin America by IDC. Clients include Lockheed Martin, GlaxoSmithKline, OfficeMax, Pemex, Cisco, Petrozuata, Luchetti and Verizon. Neoris is a privately owned company headquartered in Miami, Florida, with over 1,000 employees in the U.S., Europe and Latin America. www.neoris.com/ . Contact: Cheryl Alden, Senior Director of Marketing VoiceGenie Technologies Inc. calden@voicegenie.com 1.416.736.0905, Craig Templin, Solutions Director at Neoris

Acapela gives a voice to EDA's new Belgian directory service: Acapela Group gives a voice to EDA's 1313 service, a new arrival which adds a second directory service to the Belgian public. 1313 is a new number with 24/7 availability, providing a complete, user-friendly but efficient directory service. 1313 can give any of the listed Belgian phone numbers, and/or provide a reverse search. The service gives the phone number, but also the full address of the recipient. All the results are read out in real time by Acapela's speech synthesis, with an extremely natural and intelligible voice. Acapela and EDA have coupled their know-how to make the most of speech synthesis and adapt it to the special linguistic context of Belgium. EDA made a point of offering its Belgian users a service that takes full account of the pronunciation difficulties of related to French words in Flemish, and vice-versa, for optimal comprehension and pertinent restitution of the information. Acapela provided the know-how and the service offer that were required to make speech synthesis the key part of the 1313 service. A town like Schaerbeek, in the outskirts of Brussels, which a traditional French TTS would pronounce shar-bake, will be correctly pronounced in the Belgian-French way, i.e. skar-bek. The same applies to many Belgian surnames such as Vandenbroek, Peeters, Janssen, etc. EDA had already ruled out the traditional solution of number concatenation and preferred the smoother, more comfortable solution of speech synthesis. TTS technology also enables 1313 to add the full address to the search results, an innovation in Europe. Another key point of the service: the same voice is used to deliver all the information, whether of French of Dutch (Flemish) origin. This provides a smooth, homogenous, natural-sounding service, for better customer satisfaction. These regional concerns fully represent the quality of the 1313 service and show the expertise and added value of Acapela's linguistic team. www.1313.be www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471 Laurent Baeke, co-founder of EDA

One Voice to Offer Alternative to Directory Assistance in Texas: One Voice Technologies, Inc. (OTCBB: ONEV - News ), announced that Central Texas Telephone Cooperative, Inc. has signed a contract with One Voice to offer One Voice's MobileVoice service to Central Texas residential and business subscribers. Central Texas will first launch, to all their subscribers, One Voice's voice-enabled Alternative to Directory Assistance (ADA). This voice-enabled directory will be available from any Central Texas telephone via a short code, for example #0, and is equivalent to having the entire local phone book available from your home, business or mobile telephone. Additionally, Central Texas Telephone Cooperative will introduce other services from One Voice, such as: Voice-Dialing, Group Conference Calling, Reading and Sending E-mail, and Voice-to-Text SMS Messaging, all powered by One Voice's patented voice technology and all accessible from the local phone line. Central Texas Telephone Cooperative, Inc. was established in 1951 to provide affordable telecommunications to the people of Central Texas. When it began, the Cooperative had 614 members and provided dial tone. Today, the Cooperative services thousands of access lines throughout Central Texas and provides a wide array of telecommunications services for its members. www.onev.com dweber@onevoicetech.com

Verizon voice 411 system struggling: Darby Bailey is a 34-year-old California actress. She has three birds and likes to read short stories by Anton Chek--hov. And some people in New England do not like her. It's not that they have ever met her. They haven't and they probably never will. But they talk to her a lot. Sort of. Bailey's is the cheerful recorded voice you hear when you call Verizon Communication Inc.'s 411 directory assistance on your land line. She is the human end of a new system of speech-recognition software launched by the phone company in September and intended to make the service more efficient. But so far, the system has been struggling. More often than not, Verizon acknowledges, the system can't find the number it's been asked for and Bailey sorrowfully surrenders the call to a human operator. Although considered pretty skilled for a computer system, she sometimes can't understand what callers are saying despite multiple attempts. She apologizes a lot. She is relentlessly upbeat. And it drives some people crazy. ''Customer care has definitely decreased for Verizon with Darby," said Meg MacRae, managing partner of The Paisley Group in Castle Rock, Colo., which evaluates the nation's leading directory assistance providers. ''One big area with Darby is the calls that have to go to the operator because she can't find the number, which means customers are asked to repeat what they want three times. Customers detest being asked more than once, much less twice." Striking the right tone with customers, not to mention simply understanding them, is no easy task for a computer system. Technicians at Verizon and at TellMe Networks Inc., the California company which provides voice applications for a host of companies, have been continually adjusting Bailey's responses. At first, customers said she spoke too slowly. Bailey, who is described by TellMe as ''one of the best voice actors in the world," picked up the pace. Then, grappling with New Englanders' singular way of talking, Bailey went back into the studio and rerecorded countless business names and locations. And when callers became confused by her apologies for being unable to find a number, her mournful ''Sorry" was replaced by, ''OK, one moment" before they were whisked to a human. ''Customers thought they were doing something wrong when they heard her say, 'Sorry,' " explained Joe Horton, Verizon's executive director of Live Source technology. Bailey, who used to sing in a band called Eurydice, wound up a voice actress after leafing through scripts her husband brought home from his job as an engineer at TellMe. She admits that the job can be a strain. Recording 40,000 city names has taken months, and Bailey, who sits in a recording studio reading prompts for two hours at a time, admits it can all get a little boring. ''I try not to think about it being boring, but it can be a test of my patience," said Bailey. ''I have to keep my posture up and stay hydrated so I don't slow down. I just try to sound conversational and keep it real so I don't get syncopated. You know, I want to sound empathetic." And Verizon believes she does; it says many customers think they are talking to a human being. Not that it has been an easy transition. Horton said customer satisfaction dropped by 30 percent in some markets when Bailey was introduced but has risen steadily as improvements have been made and people become more accustomed to the system. Currently, according to Horton, 70 percent of customers surveyed rank their experience in calling 411 as very good or excellent, up from a low of 50 percent. ''One key thing we have learned is that customers really like the automation and they like it as much as the real operator," said Horton. ''Obviously, this is a work in progress, but most folks want the number quick and accurate and that is where automation hits. Darby's voice is very engaging and it sounds very pleasant." Bailey was chosen for the job because her voice ''is friendlier than your usual operator, she's the girl next door," said Jonathan Katzman, TellMe's general manager of directory assistance. But Bailey thinks it has to do with her mother. ''My mother is a music teacher and she told me to enunciate because I tend to slur," said Bailey. ''So I can really enunciate if I need to." Much of the time, however, the system cannot provide the number. According to Verizon, of the 1.6 million calls seeking government or business listings received each day, the system can take care of 30 percent itself while the other 70 percent are passed along to the human operators. Horton says that ''puts us very ahead of the game" within the industry. Overall, the Paisley group found that the system was doing a good job. Although MacRae declined to compare it's overall performance to other providers, she said that it did better by some measures. For example, it was able to understand the requested location on the first try 88.7 percent of the time. Jim Smith, a Verizon spokesman, said when the system fails, it is often as much the fault of the caller as the computer system. ''You know, garbage in, garbage out," said Smith. ''People don't have good information. They insist it's in Braintree, when it's in Weymouth. They say it's John when it's not." Either way, some callers tire of the process. Although Verizon said that less than 1 percent of callers attempt to circumvent Bailey by hitting zero as soon as the call gets underway, an unscientific survey by the Globe found some people do just that. Here's how the system works: Bailey takes all callers seeking government or business listings in the 29 states where Verizon provides local service. The requested listing is translated into speech patterns that are matched by the technology to a vast database of possibilities. While the search takes place, the caller hears a curious series of bleeps known as search tones. Verizon spokesman Smith admits they are not actual equipment sounds but ''noise intended to give the caller confidence that they are still connected and something good is happening. Dead air would be confusing." If a match is found, the system triggers Bailey to offer the listing. If not, she says, ''Sorry, I didn't get that." In most cases, she tries a second time. The system was unable to decipher a number of local addresses sought by the Globe. Asked for Worcester, she offered Norton. Asked for the Cambridge restaurant Upstairs on the Square, she first offered Upper Pendleton and then suggested something nearly unintelligible that sounded like, ''You said, you've lost your hair?" When asked for the Boston Athenaeum, she handed the call directly to a human. And that's where The Paisley Group found one of the system's biggest problems. The human operators who pick up where Bailey falls short often aren't listening in, as they are supposed to. The human operator hears a ''whisper" of a caller's request and is supposed to be ready with the listing if the caller is bounced to them. But Paisley found that operators had to ask for the listing again in 40 out of 300 test calls, resulting in an average call time of 62.3 seconds, or as MacRae put, ''a very long time." ''Customers hate that because they think the operator is not even listening to them," said MacRae. ''I won't say the reason why that happens, but there is an opportunity to mitigate customer care issues in an automated environment by working with the humans. I would imagine they do not particularly like Darby." She's right. Several operators interviewed -- none would use their names -- said that the system created more work because they had to ask the listing again. One, who identified herself as Operator F6145, said that when she used 411 herself, it could rarely understand her. ''I am not a big fan," she said. ''I think I am much better." But Bailey has her supporters, too. Megan Dyer, spokeswoman for TellMe, says that she gets several calls a month from Darby fans. They want to know what she does for fun. (Runs. Works on a screenplay.) What she looks like. (Petite. Blond.) Where she lives. (They're not telling.). ''Some people are fascinated with her," said Dyer. ''They hear her voice all the time and they want to know if she is a real person. They just want to know anything they can about Darby." Sally Jacobs can be reached at sjacobs@globe.com

Loquendo Unveils Loquendo TTS Director to Design Effective Prompts for Speech Applications: Loquendo, the global speech technology company, has released Loquendo TTS 6.4 - its latest software technology - which includes an important step forward in application development enhancement: Loquendo TTS Director. Loquendo TTS Director is a multi-platform Java development tool, which supports users in designing effective prompts for their applications. Text is written and interactively refined through a "listen & edit" procedure, which allows fine-tuning for even better TTS performance. Prompt designers can effortlessly select the TTS voice, which may be changed through control tags in the text itself; they can choose among Multi-line, Paragraph, or SSML Modes; they can set acoustic and prosodic parameters - i.e. sampling frequency and coding, pitch, speaking rate and volume -, and save their edited prompts both in text and audio formats. An easy-to-use, detailed menu helps them choose all the right TTS features. The Control Tags menu provides access to Loquendo's TTS Control Tags, which are grouped according to a set of categories - i.e. voice, language, prosody, pronunciation, spelling and pausing, audio mixer, enabling the right choice to be made easily. The Effects menu smoothly guides users through the software's advanced features, which include "expressive cues" and "plug-in lexicons" to obtain the effect required. They are grouped according to intuitive linguistic categories. The repertoire of Expressive Cues consists of a set of pre-recorded formulas, which include conventional figures of speech, like greetings and exclamations (e.g. "hello!", "oh no!", 'I'm sorry!"), interjections (e.g. "Oh!", "Well!", "Hum"..) and paralinguistic events (e.g. breath, cough, laughter, etc.), which suggest expressive intention (to confirm, doubt, exclaim, thank, etc.). This feature makes vocal messages sound even more lifelike and expressive. Loquendo technology can be further enhanced using the Plug-in lexicons, which are available to give the engine extra capability for reading specific types of text - e.g. SMS, e-mail - that may contain idiosyncratic spelling of words, abbreviations, symbols, and so on. www.loquendo.com/ Contact: Ornella Ambrois, Tel. +39 011 2913423 Ornella.Ambrois@LOQUENDO.COM

Pronexus Announces Upgrade Program for CT ADE Users: Pronexus has announced an upgrade program for users of CT ADE, Intel's NetMerge computer telephony application development environment. CT Advance entitles current CT ADE customers to a free upgrade of a development seat to the most recent version of VBVoice, Pronexus' telephony and speech app development software. Pronexus will also provide porting assistance and license discounts for existing CT ADE applications. VBVoice supports the latest telephony technologies, including VoIP, host media processing (HMP), and leading speech recognition technologies. The software uses the Visual Studio .NET environment rather than a proprietary scripting language. Pronexus says this lets developers use existing programming skills such as VB .NET to create telephony and speech applications. Developers also benefit from the familiar debugging environment and standard application extensibility that .NET offers. The free version of the CT Advance upgrade includes: VBVoice 5.2 " one developer seat, free of charge; Two free runtime licenses; and A CT ADE migration document to guide developers through the IVR upgrade process and the migration from CT ADE to VBVoice. Pronexus is also offering an extended package that includes technical support and consultation services for $1,995. Both offers require proof of current ownership of CT ADE software. www.pronexus.com Contact: Sonia Langenberg, Marketing Communications, Pronexus Inc., Tel: (613) 271 8989, sonia.langenberg@pronexus.com

Paracon Adds Pronexus Computer Telephony and Speech Development Tools to Progressive Product Offering: Pronexus Inc. and Paracon have reached a distribution agreement. Under the terms of the agreement, Paracon will distribute the full range of Pronexus products, including the market leading VBSALT and VBVoice telephony and speech software development tools. Paracon is a distributor of converged communications and computer telephony integration solutions. Paracon can now provide its channel customers with a comprehensive range of telephony and speech development tools that leverage the Microsoft .NET platform and Visual Studio .NET for maximum developer productivity. VBVoice is a highly scalable rapid application development (RAD) tool for telephony and speech. Over 3500 Pronexus resellers and customers have used VBVoice and its comprehensive range of telephony controls to rapidly create sophisticated speech-enabled interactive voice response systems (IVRs), emergency notification solutions, bill payment and call center applications, unified messaging applications, voice portals, and high volume fax solutions, to name just a few. Paracon will also distribute VBSALT, a rapid application development tool for Microsoft Speech Server. ScanSource, Inc. is a leading international distributor of specialty technology products, including automatic identification and data capture (AIDC) and point-of-sale (POS) products through its ScanSource sales unit; Avaya communications products through its Catalyst Telecom sales unit; communications products from Intel and NEC through its Paracon sales unit; and electronic security products through its ScanSource Security Distribution unit. The Company serves the North America marketplace and has an international segment, which sells AIDC and POS products in Latin America and Europe. Founded in 1992, the Company markets products from more than 80 technology manufacturers to over 15,000 value-added technology resellers and is committed to empowering them with tools and services designed to help them grow. 800.944.2432 www.scansource.com Contact: Sonia Langenberg, Marketing Communications, Pronexus Inc., Tel: (613) 271 8989, just say "Sonia" sonia.langenberg@pronexus.com Clay Sorensen, president of Paracon

VoiceObjects awarded Patent: VoiceObjects has been awarded a patent for "Dynamic creation of a conversational system from dialogue objects". This patent, granted with No. 101.47.341 by the German patent office, covers the principle of VoiceObjects’ object-oriented dialog creation, the storage of dialog design as persistent meta-data, and the dynamic generation of mark-up code e.g. VoiceXML at call time. This demonstrates to customers and partners that VoiceObjects products are based on a solid foundation.

Acapela joins Renesas COMET initiative: Renesas Technology Europe has launched COMETTM (COMmunity Enabling Telematics), the company's US-founded consortium aimed at bringing integrated systems solutions to the car infotainment and telematics market. The group helps its members to jump-start development projects, maximising the resources and time needed to implement products. COMET was formed in October 2002 to accelerate the proliferation of car infotainment system and telematics products based on SuperH(tm) (SH-4 series) RISC processors and chipsets. The consortium enables Renesas Technology and member companies to offer customers comprehensive systems solutions by utilising a readily available and optimised pool of application software. This includes Real Time Operating Systems, speech recognition, Acoustic Echo Cancellation (AEC) and noise suppression software, Bluetooth(tm) application stacks and GPS software. As a result of the COMET initiative, system engineers in automotive equipment manufacturers can now easily evaluate multiple software and middleware products running concurrently on the SDK7780 and "Biscayne" platforms for the respective SuperH-based microprocessors, SH7780 (SH-4A) and SH7760 (SH-4). Renesas Technology and COMET members have also set up and debugged key development system configurations for various telematics and CIS applications, enabling entire solutions rather than just individual software and middleware pieces to be rapidly and accurately analysed. www.renesas.com

Gionee selects ART smARTspeak XG as Speech Recognition Standard for its line of mobile phones: Gionee Corporation has selected the smARTspeak XG embedded speech interface from ART as the voice recognition standard for its new line of handheld phones. Founded in 2002, Gionee offers more than 10 phone models and has established itself as one of the leading research and development organizations in the Chinese mobile phone industry. The new Chinese version of smARTspeak XG voice interface has been integrated onto the TTPcom ADI430 chipset, which is used by Gionee for its newest line of mobile phones. The smARTspeak XG is the first embedded speaker independent name dialing solution designed for wireless phones and mobile devices. With swift and accurate recognition based on phonemic identification, the smARTspeak XG needs no training to provide instant name dialing. The smARTspeak XG is fully speaker independent, meaning it can be operated by any user with no voice training of the device. In addition, it incorporates continuous digit dialing, speaker independent command and control, and trainable continuous digit dialing for custom languages. The advanced features of the smARTspeak XG enables users to dial by speaking a person’s name or phone number. Giving equally fast and accurate response whether users say "Xiao Wei" or "Dial 1202111002", smARTspeak XG maintains ART’s popular out of-the-box, multi-user experience. In addition, the phone will recognize commands such as "contact" and "redial" and offers trainable digit dialing that can be adjusted to a user’s accent. When more than one name is available, ART’s "Best Match" function presents a short-list of options either as text or voice for extra convenience. Today, ART's embedded software-only solutions are deployed in dozens of product lines from industry leaders such as LGE, Motorola, Mitsubishi, NEC, Panasonic, BenQ, Xelibri by Siemens, Sierra Wireless, Compal, Quanta, Inventec, Gionee, Siemens Automotive, Maxon, GVC, Hitachi, Emblaze mobile, Casio, Franklin, PaceBlade and others. ART has strategic and technological partnerships with industry leaders including: Texas Instruments, Qualcomm CDMA Technologies, Freescale, TTPCom, Agere, Infineon, Intel, Broadcom, ADI, Microsoft, Symbian, UIQ, Series 60, Prairiecomm, Spreadtrum, Renesas, Cirque, Accelent Systems, Teleca, Ceva, Clarity, e-sim, Persay, SVOX, DSPG, Pegasus, VOXON and others. www.artcomp.com Established in 2002, Gionee Corporation develops and manufactures mobile communication devices primarily for the Chinese domestic market. With headquarters in Shenzhen, China, the company employs a staff of more than 1,600. The company hs established the Gionee Mobile Communication Design Institute, which has quickly become one of the strongest research and development groups in the domestic field. http://www.gionee.net Contacts: Betty Finkelstein, ART, Ltd., betty@artcomp.com +972-3-7685018 Zhang gaoxian, the IMC Director of Gionee

NEC America Selects ART voice control for its new NEC 232E high definition mobile handsets ART Advanced Recognition Technologies announced that NEC America, Inc. (NEC), a leading provider of innovative communications products, solutions and services has selected the smARTspeak NG embedded speech interface for its recently launched NEC 232E+ High Definition Mobile++ (HDM?) handset. With no advanced training necessary for the speaker-independent voice interface, NEC customers can enjoy the phone’s voice control features right "out-of-the-box." www.artcomp.com Betty Finkelstein Marketing Communications ART, Advanced Recognition Technologies Ltd. Tel: +972 3 7685018, Mobile: +972 524 700018 betty@artcomp.com

Fonix VoiceDial Now Shipping on T-Mobile and O2 Microsoft PocketPC Devices: Fonix Corporation (OTC BB: FNIX), an integrated communications carrier providing telecommunications services and value-added speech technologies, has licensed its Fonix VoiceDial hands-free dialing technology for use by High Tech Computer Corporation (HTC), an original design manufacturer (ODM). The Fonix/HTC agreement is based on unit royalties. Current mobile operators marketing HTC devices in Europe featuring Fonix VoiceDial include O2 and Deutsche Telekom’s mobile communications division T-Mobile. HTC selected Fonix VoiceDial to be a standard software feature on its devices marketed to global mobile operators and original equipment manufacturers (OEMs). New to the market are the T-Mobile MDA III and the O2 XDA II PocketPC devices, both featuring Fonix VoiceDial in multiple languages, which allows users to access their mobile phone contacts simply by speaking, with no voice training required. To purchase the T-Mobile MDA III, please visit T-Mobile Germany at www.t-mobile.de/index . To purchase the O2 XDA II, please visit O2 Germany at www.de.o2.com/static/home/flash7.html www.fonix.com HTC is the world’s largest original design manufacturer (ODM) of windows CE-based Pocket PC, Wireless Pocket PC, and Smartphone products. www.htc.com.tw www.fonix.com Contact: Elizabeth Sweeten (801) 553-6600 mediainfo@fonix.com

Vodafone Expands Launch of Speaking Phone to Speech-Enable Handsets for Blind and Low-Vision Customers ScanSoft ScanSoft, Inc. (Nasdaq: SSFT) announced that Vodafone is expanding its offering of ScanSoft SpeechPAK TALKS application for blind and low-vision customers across Europe. Already widely used in several countries around the world, the Vodafone Speaking Phone is based on ScanSoft's SpeechPAK TALKS application, and provides blind and low vision consumers access to many of today's advanced wireless services, including e-mail, short messages, menus and contacts. The Vodafone Speaking Phone is available on a range of Symbian Series 60 handsets in the United Kingdom, and South Africa, and it is intended to be introduced in a further six countries during 2005, including Ireland, Italy, Greece and Portugal. According to the World Health Organization, there are more than 180 million people who are blind or have low vision, and the visually impaired population has expanded more than 50% in the last decade. Vodafone Speaking Phone directly addresses this expanding population by providing a full-featured mobile phone that enables individuals to have access to information and common features on their mobile phones - from short text messages to caller ID -- read aloud to them in a natural-sounding voice. Vodafone Speaking Phone provides users with the information they need quickly, in a comfortable, easy-to-use manner, alerting them to messages such as email, text messages and incoming calls, and provides a unique caller ID function that enables users to learn who is calling without picking up the phone. Typical caller ID features are visually based, and force the blind or low-vision caller to answer the phone to learn who is calling. With Vodafone Speaking Phone, the caller's identity is read aloud to the user in a voice that is configurable and easy to understand. In addition, speech-enabled access to practical features, such as battery and signal strength, ensures that users can maintain the basic functions of the telephone. Based on ScanSoft SpeechPAK TALKS, Vodafone Speaking Phone provides a host of features designed to make mobile phones more accessible, and to directly meet the needs of the blind and low-vision community or provide accessibility to those who simply wish to become more productive through hands-free access to important information. With Vodafone Speaking Phone, users can: Hear caller ID information and check logs of incoming and outgoing calls *Write and listen to 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, alarm clock and other important tools * Hear practical information about the battery level, network and signal strength . www.ScanSoft.com . Vodafone is the world's largest mobile community, with equity interests in 26 countries and Partner Networks in a further 13 countries, serving over 147 million proportionate customers. Guy Laurence, Global Marketing Director Consumer, Vodafone

Royal National Institute of the Blind (RNIB) Endorses ScanSoft SpeechPAK TALKS: ScanSoft, Inc. (Nasdaq: SSFT) and the RNIB the UK's leading charity offering information, support and advice to around two million people with sight problems, announced that the RNIB has signed an extended agreement to offer ScanSoft SpeechPAK TALKS to individuals who are blind or visually impaired throughout the United Kingdom. There are around 2 million people in the UK with sight problems and every day in the UK another 100 people start to lose their sight. SpeechPAK TALKS directly addresses the growing concern by providing a "talking phone" that enables individuals to have full access to features and information on their mobile phones - from short text messages to caller ID read aloud to them in a natural-sounding voice. Leading service providers, including Cingular, Vodafone, Orange and Telecom Italia, and a global network of assistive technology dealers currently market and offer SpeechPAK TALKS to customers across the globe. RNIB's objective is to ensure that mainstream distribution channels are utilised to deliver mobile technology to visually impaired and blind users. It will concentrate on business to business relationships with the mobile industry in order to make sure customers get a good service, a good price and the product they want when they want it. Over the next two months agreements are being put in place alongside ones that already exist such as RNIB's partnership with Vodafone. www.ScanSoft.com/speech . www.ScanSoft.com . RNIB's pioneering work helps anyone with a sight problem - not just with braille, Talking Books and computer training, but with imaginative and practical solutions to everyday challenges. We fight for equal rights for people with sight problems. We fund pioneering research into preventing and treating eye disease. We need your generosity. If you or someone you know has a sight problem, RNIB can help. Call the RNIB Helpline on 0845 766 9999 www.rnib.org.uk Steve Tyler, Strategic Manager Digital Technology, RNIB

Japanese Robot to Chat with Elderly: Japan's growing elderly population will be able to buy companionship in the form of a 45-centimeter (18-inch) robot, programmed to provide small talk. Snuggling Ifbot, who is dressed in an astronaut suit with a glowing face, has the conversation ability of a five year old, the language level needed to stimulate the brains of senior citizens, its software designer said. If a person tells Snuggling Ifbot, "I'm bored today," the robot might respond, "Are you bored? What do you want to do?" To a statement, "Isn't it nice today?" the robot could say, "It is a fine autumn day," by detecting the season from its internal clock. The company has received 128 pre-launch orders for the robot which hit stores Wednesday at a cost of 576,000 yen (5,600 U.S. dollars). Other than the conversation function, the robot has 15 programs to keep the elderly thinking and healthy, including singing songs, reading out quiz games and old news, and inquiring about health functions. While for now Snuggling Ifbot speaks only Japanese, its makers plan to program the robot in English -- not for export, but to teach the language to Japanese children. The Japanese are famous for their longevity, with more than 23,000 people aged 100 or over, in part due to a traditionally healthy diet and active lifestyle. Japanese society is set to be getting ever grayer as more young people put off marriage and decide against having children. Japan's birth rate hit an all time low of 1.29 children per woman in 2003. The robot was developed in cooperation with Business Design Laboratory, Futaba Industrial Co., and Brother Industries.

Sensory Launches Ultra-compact Speech Recognition technology for embedded applications: Sensory, Inc. (Santa Clara, CA) debuted FluentChip technology for the RSC line of microcontrollers. FluentSoft technology is available to be licensed to other manufacturers for use on multiple platforms. FluentChip and FluentSoft came to Sensory from the acquisition of Fluent Speech Technologies, a group of academics out of the Oregon Graduate Institute’s (OGI) Center for Speech and Language Understanding. The new technology will be demonstrated at CES in Las Vegas, January 6-9, in Sensory’s meeting room #36079. Dr. Pieter Vermeulen, a former OGI professor, and now Sensory’s director of technology, joined Sensory when the technology was acquired. The Fluent team, led by Dr. Vermeulen on the FluentSoft side and Dr. Robert Savoie on the FluentChip side, have spent over four years perfecting a technology that requires as little as 20K of memory to recognize a word set with extremely high accuracy, speaker independence and reasonable tolerance for ambient noise. Sensory’s RSC line of microcontrollers, with speech I/O capabilities, now run the FluentChip software. These low-cost ICs selling for under $2 in volume, can handle active vocabularies of up to 40 speaker-independent words. All the necessary ROM, RAM and other hardware are contained on the chip, including a general-purpose microcontroller and powerful 16-bit A/D converter. By replacing an existing non-speech microcontroller, Sensory can add speech input and output, with an incremental cost of under $1. This opens up the technology to a new generation of consumer products, with user interface features that were never before affordable. Fast, Easy Development Tools Cut Development Costs by Orders of Magnitude New language development tool kits for FluentChip cut speech development time from months to hours. Programming tools like C-compilers, debuggers and emulators may be used to speed programming time. FluentSoft is compatible with a number of popular DSP’s, microcontrollers and various operating systems. FluentSoft works well on a cell phone platform recognizing over 600 names and continuous digits, even in a noisy environment. Sensory maintains a focus on improving the user interface in consumer electronics through speech technologies. Sensory sells both IC and embedded software solutions and has received venture capital financing from Norwest Venture Partners, MT Partners, Convergent Capital, Uniden Corp., Mitsubishi Ventures and Investar Capital. The company’s ICs are widely used in consumer electronics applications including telephones, home automation, toys, remote controls, automotive, security, learning aids, and other markets where speech input and output is desirable. Sensory’s customers represent the leaders in consumer electronics, including such companies as Hasbro, JVC, Kenwood, Mattel, Mitsubishi, Toshiba, Uniden, Sony and more. Sensory has other offices in Portland, Oregon; Hong Kong; Tokyo and Vienna. www.sensoryinc.com Contacts: Sensory, Inc. Tom Tolbert 408/240-1577 ttolbert@sensoryinc.com KPR, Inc. David Kaye 818/368-8212 dave@kprinc.com

Wizzard Software begins Shipping Talking Pill Bottle Starter Kits to customers: Wizzard Software (OTCBB: WIZD) and its subsidiary, MedivoxRx Technologies has begun fulfilling orders and distributing its new Rex the Talking Bottle Starter Kit for personal use. The Starter Kit, formerly known as the Home Kit, allows the patient or caregiver to record customized, verbal medication instructions. The Starter Kit, which is attractively packaged for retail shelves, contains easy-to-follow instructions, power supply, 3 bottles and a microphone that makes the recording possible by simply pressing a button and speaking. A caregiver for instance, can record "patient friendly" instructions for over-the-counter or prescription medications as well as refill reminders and medication warnings. The medication instructions can then be heard at any time by the patient by pressing a button on the side of the bottle. This is particularly useful for the elderly who may be confused by complex technical terms and instructions in addition to those with vision problems that make reading a traditional label difficult. After using the initial three bottles in the Starter Kit customers can purchase a Refill Kit containing three additional bottles. In addition to shipping product to traditional pharmacies, blind rehabilitation organizations, consultant pharmacists and senior care facilities have expressed strong interest in the Starter and Refill kits for their clients and patients. These products are also available on the GSA Advantage web site at www.gsaadvantage.gov (keyword Wizzard). www.rxtalks.com The products can be purchased today at www.rxtalks.com/purchase/index.htm 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 PR Contact: KPR, Inc. Dave Kaye 818/368-8212 dave@kprinc.com

Readspeaker's voice-on-the-web services spread on the web: Readspeaker, a leader in voice-on-the-web services in Europe, is equipping more and more websites with Acapela's text-to-speech technology. The latest websites that are now speaking out are the district of Grand Chalon, and Terre-net, the first French agricultural news website, with over 2 million pages viewed per month. The district of Grand Chalon ( http://www.legrandchalon.fr ), in the heart of the Bourgogne region, counts 108,000 inhabitants over 38 villages and towns. The district decided to speech-enable its website to make it more easily accessible to all as part of a global project. This precious service is based on a simple figure : 10% of the French population, roughly 6 million people, have sight-related problems. Beyond the concern of the visually impaired, Readspeaker also allows users to overcome obstacles such as illiteracy or language problems, and can also just make the interface more comfortable for the average user. Terre-Net ( http://www.terre-net.fr ) had the same concerns when they decided to ease the access to the agricultural news on its site. Terre-Net is the first French portal to offer speech-enabled access to its content. Today, Terre-Net allows its visitors to listen, at the click of a button, to contents that can be updated up to 20 times a day. This is made possible by the speech synthesis which can read any written information in real time. "By vocalising the news articles on our site, we intend to show our engagement towards an information system that is accessible to the visually impaired." says Arnaud Romoli, CTO of Pole Terre-net Médias. Readspeaker technology is a service available on the internet, even with a low-speed connection. It requires no installation, neither on the users computer, nor on the the site's server. The information is read aloud in real time with a intelligible and natural voice, provided by Acapela's speech synthesis. Grand Chalon district and Terre-Net join GlaxoSmithKlein laboratories, Sesam Vitale, Yanous! and the French National Assembly (on test until end of 2004) who have already adopted Readspeaker's solution. www.readspeaker.fr www.acapela_group.com Contact: Caroline Houël chouel@acapela-group +33 (0)5 62 2471

Acapela gives Mona Lisa the say in a leading Paris museum. The "Cité des Sciences et de l'Industrie" - a key institution in scientific, technical and industrial cultural fields and 4th most visited museum in France with over three million visitors each year - has just unveiled a new attraction dedicated to "Sounds". At the heart of thirty-one different animations looking at sound from every angle, Acapela, long-time partner of the Cité des Sciences for the demonstration and promotion of the use of vocal solutions, opens up a dialog between Mona Lisa and the visitor. Mona Lisa, transformed into a virtual character, calls out to the visitors, inviting them to play to a game and test their knowledge on the exhibition theme, "Voice and Hearing". The visitor can also chat with Mona Lisa, making compliments or comments, or ask questions about her. "This interactive stand allows us to present and apply speaker independent Automatic Speech Recognition. The visitors, from kids to adults, can recap the knowledge acquired thanks to the exhibition while still playing with sound, in this case driving the interface with their own voice, with very natural interactivity. This playful stand perfectly fits in the mission of the Cité des Sciences to make cultural and technical know-how known to the general public." states Françoise Vallas, Manager of the Sounds exhibition, at the Cité des Sciences et de l'Industrie. Acapela gives the Cité's visitors the best of its speech technologies to fully demonstrate an intuitive, convivial and natural dialog interface www.cite-sciences.fr/english/indexFLASH.htm Explora exhibitions program.

Animated & speech enabled characters by Inovani. Inovani, a leading provider of speech animation products has selected Acapela's high end TTS to voice enable its animation engine, thus providing "talking heads" with a very natural and smooth animation and personality result. Inovani delivers an animation engine: a software only, state of the art rendering of talking expressive characters. These characters are programmed to present spoken content with close to natural facial motions, and they include life-like human head models, suitably modeled cartoon heads, "talking heads" or avatars. "Associated with text to speech, the character gallery comes to life. Acapela text to speech high quality and naturalness fits perfectly with the realistic speech motions of Inovani" said Else Nordhagen, Inovani CTO. Speech motion is so realistically defined that it allows lip reading. It can be combined with emotions such as smiling while talking! Just try the demo: all you need to do is type your text, add the smilies (happy, surprised, secretive, interested, questioning, raise eyebrows, etc.), choose your character, look & listen! Inovani enables Telecom operators, Content providers and others to offer value added services, on both mobile devices and PCs, by offering platform independent products for interactive life-like animation of expressive talking characters. Acapela TTS reads aloud any written information, in real time, opening up a wide panorama of personalised and catching services. For Christmas time, Inovani offers to create your own greetings using the Acapela voices at http://www.facems.com The technology is language independent. The animation software runs on PCs and PDAs and in distributed environments. Inovani's products can be integrated with various applications: Web browsers, MMS services, games or specially targeted applications. www.inovani.com

Interactive Multimedia does Vampire Lip Sync: Interactive Multimedia Solutions, Inc. (Apex, NC), a provider of voice-to-animation solutions for the multimedia market, announced that Activision, Inc. (Nasdaq: ATVI) and Troika Games licensed the IMS V2A MDK to assist in the lip-synchronization and facial animation sequences in the newly released Vampire: The Masquerade – Bloodlines. Currently available for the PC at retail outlets nationwide, Bloodlines is an all-new RPG adventure driven by a devious and intriguing storyline, where gamers become creatures of the night, wandering the dark and gritty vampire world of modern day L.A. while a mysterious plot reveals itself. The IMS V2A MDKTM is a file-based analysis engine that takes as input a human recorded voice file, or one generated from a text-to-speech (TTS) engine, and automatically creates precise mouth, jaw, and lip position data. The audio files can be in any language. The IMS voice-to-animation engine is ideal for English, but because of its unique proprietary co-articulation algorithm, quality animation data can be generated for French, Italian, German, Spanish, Japanese, Mandarin, and a number of other languages. This is a significant time and cost-savings advantage to game developers that want to localize their game titles for multiple markets. The IMS V2A MDK also produces corresponding automated speech gestures that are generally associated with speaking including head, eye, eyebrow movement, and eye blinks. These automated speech gestures in combination with the lip synchronization data can be use to automatically animate secondary game characters without the need for additional tweaking. Additionally, the product's flexible API supports the use of text in combination with audio files to generate more precise lip synchronization data, and makes it a must-have for serious game developers and game publishers. The IMS V2A MDKTM product is licensed to game developers on a per/title basis. www.IMS3D.com Contact: 919-389-6430 Donovan Moxey, CEO demoxey@ims3d.com

MacSpeech Announces New Transcription Professional Solution: Macintosh speech recognition authority MacSpeech announces the immediate availability of its new Transcription Professional Solution featuring the Olympus DS-4000 Digital Voice Recorder. This solution provides the greatest ease of use of any Transcription Solution, yet will meet the needs of the most demanding professional with features like a four function side slide switch, more than 11 hours of recording time, automatic backup of recordings, and multiple user profiles to support environments where recorders are exchanged amongst professionals. The iListen Transcription Solution Professional 4000 sells for $549.00 including the DS-4000 recorder, iListen software, and a high-quality VXI-Pro headset. It is available immediately from the MacSpeech web site. www.macspeech.com Contact: Chuck Rogers (504) 628-3640 chuck.rogers@macspeech.com

Opera Tackles Voice Browsing and RSS in Latest Beta Release: Opera Software ASA (Oslo, Norway) unveiled a beta test version of its next Web browser release that features speech recognition, discovery of news feeds and automatic. Opera plans to make its newest browser generally available in 2005. While Opera supports Windows, Mac OS X and Linux, the beta is available for Windows only. Opera is the biggest commercial competitor to Microsoft Corp.'s Internet Explorer, though it falls behind open-source contender, Mozilla, in user share. Opera holds about a 1.3 percent share worldwide, compared to 88.9 percent for IE and 7.4 percent for Firefox and Mozilla's other browsers, according to Web analytics provider OneStat.com. With this week's beta release, Opera launched a series of voice-browsing features that the company had promised earlier in the year. Opera is working with IBM's ViaVoice technology group to create a browser that can recognize spoken commands and read Web-page content. For example, the beta lets users navigate Web sites by speaking such commands as "Opera back" or Opera next link," Opera announced. By saying "Opera speak," the browser will read content and e-mail messages. Opera ships its browser with a built-in e-mail client. In the Opera 7.50 release, the client added a news reader for RSS (Really Simple Syndication) feeds. Opera's new voice technology is said to support websites that offer interactive, voice-enabled shopping and booking systems, the company says. Additionally, users can browse the web using spoken commands such as "Opera next link," "Opera back," and "Opera speak." In response to the latter, Opera reads the content of a web page or email aloud. Opera and IBM have partnered on XHTML+Voice (X+V) technology for several years, co-announcing a Multimodal Browser and Toolkit early in 2003. At that time, Opera began offering an enhancement to Opera 7.0 browser for Windows that was made available through the Multimodal Browser project. The modified version of Opera 7.0 was capable of rendering Web pages written in the XHTML+Voice markup language . www.operasoftware.com Jon von Tetzchner, Opera's chief executive officer Jon@operasoftware.com Tel: +47 24 16 40 00

Right Seat Software’s Vox Proxy Version 2 Friendly Persuasion: Have you ever SEEN a PowerPoint presentation you could describe as entertaining? Probably not. Most likely that’s due to limited funds. With a bigger budget, we could add a professional narrator, tightly integrated music and carefully choreographed graphics. There are, however, ways to liven a presentation up without having to invest lots of time or money. Vox Proxy Version 2 from Right Seat Software adds talking 3D-animated characters to your PowerPoint slides. It’s easy to use and relatively inexpensive ($199). While many multimedia add-on products fail because the resulting quality is less than professional, the results from Vox Proxy can be quite good, depending on how you use it. The program has a built-in text-to-speech engine that sounds just as you would expect—overly electronic to be convincing. However, if you purchase Vox Proxy’s AT&T Natural Voices add-on pack ($60 for two voices), the text-to-speech voices are almost indistinguishable from real voices. The completed Vox Proxy character animations are embedded into the PowerPoint presentation and can be played back on a PowerPoint-equipped PC using the free Vox Proxy Player. You can also save your presentation to a standalone CD using the company’s CD Prep utility (it’s included in Vox Proxy Version 2 with CD Prep, which sells for $228.95). There are no royalties involved with creating a CD. You could create a self-starting training program or product demo narrated by a group of 3D-animated characters, and then distribute the presentation on CD to your employees or customers.

Sakhr Solutions and Technologies at the Microsoft Solutions Summit in Cairo: Sakhr Software Co. recently participated in the meeting focusing on technologies based on Microsoft platforms. Sakhr's participation came in the form of two workshops held on the final day of the meeting hosted by Microsoft which was held over a period of three days, from 20-22 December 2004 in the Egyptian capital Cairo, and was attended by major Arab IT companies. During the workshops Sakhr demonstrated its two highly specialized systems aimed at Arab e-Governments and the comprehensive management of educational e-portals. Sakhr presented its E-Portal system for electronic education. The E-Portal system is a comprehensive advanced system that fulfills requirements on a variety of levels. These include the management of educational regions, the management of schools sectors, the automation of school administration – teacher/ student/ parent relations, the interaction between all involved in the education process – executed via the Internet. Sakhr also demonstrated its solutions targeting Arab e-Governments which include content management, advanced search engine, automatic reader OCR to convert traditional documents into electronic files, text mining tools that include spell checker, automatic categorizer, key word extractor and automatic summarizer. Technologies demonstrated include Sakhr's speech engines which include Text-To-Speech and ASR (Automatic speech recognition) technologies that convert speech into text, and numerous other systems which integrate with the Arabic comprehensive information management solution AIMS. Sakhr develops software specifically tailored to the needs of the Arabic market and Arabic language users. It has developed cutting-edge technologies that have had a strong impact on the Arabic IT industry. The company has also pioneered a new generation of Arabic Natural Language Processing (NLP) technologies. Sakhr opened its branch in the UAAE in 2001, at the Dubai Internet City. Sakhr's list of clients includes giant IT companies such as Intel, Oracle, LG, Tejari.com & other major international & regional companies, governments, ministries & central banks. Contact the Media Department at Tel: +971 4 391 1555 Speech@sakhr.com

Vianeta Adds Speech Recognition to Harmony Health Information Management Platform ScanSoft, Inc. (Nasdaq: SSFT) and Vianeta Communications, a leading provider of Web-based XML health information management (HIM) solutions, announced the integration of ScanSoft Dragon NaturallySpeaking with the Vianeta Harmony Health Information Management (HIM) Platform. Dragon NaturallySpeaking enables the automatic conversion of speech into text at up to 160 words per minute, and allows users to control software applications simply by speaking. Vianeta Harmony is a unified e-HIM platform used by hundreds of care facilities and more than 64,000 care providers in North America to automate their patient records processes. It provides everything needed for digital dictation, speech recognition, transcription, document scanning and chart distribution, and connects with department-specific systems. Vianeta used the ScanSoft Dragon NaturallySpeaking SDK (software developer's kit) to add front- and back-end speech recognition to several components of Vianeta Harmony, including the Vianeta Speech Recognition Desktop (SRD) for Radiology, and the Vianeta Speech Recognition Enterprise Server (SRES). Dragon NaturallySpeaking allows the Vianeta Speech Recognition Desktop (SRD) for Radiology to transform instantly the physician's voice dictation into digital text, where it can be easily reviewed, edited and approved by the physician. By enabling front-end speech recognition the system speeds data entry into the electronic medical records (EMR) system, completely avoiding the time and costs associated with traditional manual transcription processing. ScanSoft Dragon NaturallySpeaking SDK is integrated in Vianeta Speech Recognition Enterprise Server (SRES) to enable the back-end batch-processing of recorded speech from a range of devices, including digital recorders, PDA's, Tablet PC's and telephone dictation systems. After Vianeta's Workflow Manager receives the dictation, Dragon NaturallySpeaking SDK transcribes dictation into text. Using the physician speech profiles that are resident in the Dragon NaturallySpeaking server, the resulting transcription is then sent to a medical transcriptionist for further editing via the Vianeta Transcription Desktop. Vianeta Speech Recognition Enterprise Server (SRES) with integrated speech recognition from ScanSoft Dragon NaturallySpeaking SDK, reduces transcription processing costs by 50 percent or more. Built on an XML Web-based platform, Vianeta Harmony is designed to uniquely address many of the shortcomings that exist with legacy clinical documentation solutions and enable hospitals and healthcare systems to leverage clinical documentation to lower costs and improve patient care. From dictation through reimbursement, Vianeta delivers standardized and cost-effective applications for the healthcare enterprise. 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 that can be accessed securely by clinicians, transcriptionists, coders, billing analysts, patients and more. www.vianeta.com . Vianeta Corporate Contact: Kelly Colburn, 408-519-2142, kcolburn@vianeta.com www.ScanSoft.com . ScanSoft, Inc. Erica Hill, 978-977-8466 erica.hill@scansoft.com

Transcend Introduces Speech Recognition Functionality: Transcend Services, Inc. (Nasdaq SmallCap: TRCR) has completed the integration of speech recognition technology into its proprietary, Internet-based, medical transcription technology platform. The speech recognition technology was provided to Transcend by MultiModal Technologies, Inc. ("M*Modal") under a seven-year, non-exclusive license. The proprietary M*Modal technology consists of a natural language understanding processor, a conversational speech recognition engine and editing tools, all of which were designed to work in concert with each other to continually learn and translate voice records into editable, structured clinical documentation with improving levels of accuracy as more voice records are processed over time. Transcend integrated M*Modal's technology into its web-enabled voice and data distribution technology. Transcend intends to apply this new speech recognition functionality in the delivery of electronic medical record documents for a pre-determined sequence of medical specialties or clinical departments. Transcend recently started utilizing the M*Modal technology to edit reports for a select number of customers. Transcend intends to aggressively increase the utilization of the M*Modal technology and expects to be using this technology to produce the majority of its customers' clinical documentation by the end of 2005. Transcend believes that accurate, reliable and timely transcription creates the foundation for the patient medical record. To this end, the Company has created Internet-based voice-to-text systems that allow its skilled medical language specialists to securely and quickly produce the highest quality medical documents. The Company's wide range of transcription services encompass everything needed to securely receive, type, format and distribute electronic copies of physician-dictated medical documents, from overflow projects to complete transcription outsourcing and custom data-center creation packages. www.transcendservices.com . MultiModal Technologies offers hosted conversational documentation services that help healthcare providers to capture discrete clinical information from dictation to generate complete and timely electronic medical records. The company's unique speech recognition and understanding technology platform, AnyModal CDS, is a vital tool that empowers physicians to capture clinical facts and orders from dictation without requiring any change to their normal dictation routine. M*Modal's focus is on providing hospitals and medical transcription service providers with the industry's most comprehensive yet most adaptable solution for creating highly accurate, structured, encoded and shareable medical documents to increase patient safety, promote continuity of patient care, and reduce cost. 412.578.9700 www.mmodal.com . info@mmodal.com Transcend Services, Inc., Atlanta, Larry Gerdes, 404-364-8000 larry.gerdes@trcr.com

Asia Broadband Inc. to Incorporate Text-To-Speech Technology from Wizzard Software in Multimedia Applications: Wizzard Software (OTCBB:WIZD) announced that Asia Broadband Inc. has licensed AT&T's Natural Voices speech engine to use in their multimedia interactive educational and entertainment applications. With operational offices located in the United States, Canada, United Kingdom and mainland China, Asia Broadband's business focus includes: investing, acquiring and controlling e-learning and online education related assets in Asia; producing and distributing educational content and e-learning software products under the IE21 brand worldwide; transferring and localizing e-learning platforms and solutions to the Asian region; and developing educational programs in partnership with both international and local educational groups in Asia.
The integration of AT&T's Natural Voices into Asia Broadband's product line will enable new learning methods and enhance the interactive experience of the user/student with new and effective feedback mechanisms. Their unique application is also designed for special needs such as for blind individuals using brail keyboards. Initially targeted for implementation into classrooms to enhance the teaching experience, Asia Broadband is currently researching other applications and environments in which they can incorporate Text-To-Speech. Asia Broadband Inc. (Pink Sheets:ASAB) is wholly focused on becoming the leading provider of online educational programs in China and throughout Asia. Asia Broadband operates in China through its 95% owned and controlled Chinese subsidiary Shanghai Broadband Network, Inc. www.rxtalks.com . Lana V. Ilchenko, Business Development Executive for Asia Broadband

AWG experience with speech recognition: Speech recognition technology is expected to explode at warehouse and supply-chain operations as companies look for new ways to cut costs and improve productivity. Propelling the trend are the stories from early adopters such as Kansas City, Kan.-based Associated Wholesale Grocers, which gave local members of the Council of Logistics Management a firsthand look at the technology on Dec. 9. AWG, a member-owned wholesaler, began using the technology for order-picking last year. In areas where it swapped scanning technology for belt-worn Talkman computers from Pittsburgh-based Vocollect, AWG officials said, the company has increased its order selectors' productivity by 8 percent in its produce area and 15 percent in dairy. In the dry goods and freezer picking areas, where AWG switched from paper to voice, productivity gains were 3 percent and 4 percent, respectively. Equally important has been the corresponding increase in order accuracy, to 99.64 percent from 99.52 percent. "If you think about the millions of cases we ship out of here on a yearly basis, a very small percentage of errors adds up to a big amount of money in a short period of time," said Rich Vastine, AWG's director of industrial engineering. To use the system, order selectors have to spend about 20 minutes one time to create a voice template. The system asks the selectors to provide short responses, such as "ready," and numerical check codes that ensure they're picking from the correct slots. When asked to repeat the number of items to be picked, the selector can advise that an item is in short supply or "out," allowing slot replenishers to kick in to gear. In addition, the pace of the system can be adjusted by the selectors. It also allows managers to monitor workers' productivity via a real-time computer program. "It's Big Brother at its best," Vastine said. Switching to the voice technology, Vastine said, means order selectors can work at their maximum pace, unencumbered by clipboards, labels or scanning devices that fog up in AWG's huge freezer. Another advantage of the new technology is that it operates with existing RF (radio frequency) warehouse-management systems. In addition, said Larry Sweeney, vice president of product management for Vocollect, the cost savings it provides will help companies find the cash to upgrade to RFID tracking technology, which Wal-Mart Stores Inc. has directed its top 100 suppliers to implement by Jan. 1. AWG did not disclose how much its system cost. But in a written release, AWG's CFO, Robert Walker, said that the company would save millions and that it had achieved a return on investment much sooner than the 18 months it had projected. In addition to saving $250,000 on bar-code labels at the Kansas City, Kan., facility, AWG has shaved the costs of workers' comp www.vocollect.com Vocollect, Brande Juart, 412-829-8145, Ext. 2448 bjuart@vocollect.com

Meet Saskia, the new Dutch Female Voice! Loquendo Male and Female Voices Set the Industry Standard for Voice Expressiveness, Quality and Intelligibility Loquendo, the global speech technology company, is pleased to announce Saskia, the new bubbly Dutch female voice to accompany Willem, the Dutch male persona. Loquendo's multilingual portfolio of lifelike synthetic speech is rich as never before. Many customers from numerous sectors, such as telecommunications, banking, voice portals and content providers are currently using Loquendo TTS. Also suited for e-mail, personal assistants, talking heads, unified messaging, mobile and CRM applications, Loquendo TTS is available in 16 different languages, in both male and female voices, and runs on all relevant operating systems. To listen to the synthetic voices of Saskia and Willem and their expressivities, please click here . To meet the Loquendo TTS family, please click here . With over 30 years of 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 ASR - automatic speech recognition, Loquendo Speech Server, Loquendo Speaker Verification) and platforms guarantee systems integrators the best solutions in 16 languages and 36 voices - with more coming soon. www.loquendo.com Contact: Ornella Ambrois, Tel. +39 011 2913423 Ornella.Ambrois@LOQUENDO.COM

ScanSoft Unveils RealSpeak Telecom 4.0: ScanSoft, Inc., (Nasdaq: SSFT) announced the worldwide release of RealSpeak Telecom 4.0 - the premier text-to-speech (TTS) solution for conversational call center and carrier-based applications worldwide. RealSpeak Telecom 4.0 turns any ordinary text - such as text from a database, directory assistance service, news report, or short message service (SMS) - into high-quality speech output. RealSpeak Telecom 4.0 delivers the most comprehensive range of languages, dialects and voices worldwide, achieving a new level of expressive, natural voice quality to enhance customer service in applications across the globe. RealSpeak Telecom 4.0 is the first product to leverage and build upon the combined assets of RealSpeak 3.5 and Speechify 3.0, independently considered leading products before this release. RealSpeak Telecom 4.0 introduces 10 new voices, enabling businesses to converse with their customers through an incredible degree of accuracy and consistency in multiple languages, including the world's first TTS solution for the rapidly growing Indian market. RealSpeak Telecom 4.0 is currently available in 22 high-quality languages and 30 superb voices with a further voice and language roll-out in 2005. RealSpeak Telecom 4.0 supports tuning tools that enable developers to optimize, tune and vary TTS prompts. These tuning and personalization opportunities allow developers to significantly enhance caller-experience in their conversational applications and to improve ROI by increasing task automation. Furthermore, RealSpeak Telecom 4.0's new voices boast sophisticated bilingual capabilities, allowing a single voice to elegantly handle words from multiple languages within the same application without sacrificing quality or pronunciation - a necessary capability in today's global economy. As business-to-customer interfaces become increasingly critical differentiators, more organizations look for unique and creative ways to strengthen their brand equity with custom voices on RealSpeak Telecom 4.0. SpeechWorks Solutions offers an unparalleled blend of experience, technologies and tools that enable the creation of compelling, effective and unique custom voices that augment call-center and carrier applications. The addition of Rhetorical's portfolio of specialized tuning tools and best practices will enable RealSpeak Telecom 4.0 customers to deploy custom voices more easily. To date, more than 15 custom applications have been implemented by the combined organization world wide, including varetis AG and Belgacom. "We are looking forward to offering our customers a new level of service that we simply can't achieve with pre-recorded prompts." said Gianluca Ferranti, Marketing Director of Reitek, a leading Italian supplier of Customer Interaction Management Solutions. "Thanks to the quality level attained with RealSpeak Silvia, for the first time, we are now considering offering text-to-speech as a standard module for our client's application." "The new RealSpeak Virginie voice from ScanSoft has set a new quality benchmark for the French market," said Christophe Joubert, managing director, Sintel. "We can't believe how consistent, expressive and pleasant the voice is, and how positively our customers have reacted to it." "ScanSoft's has made yet another amazing leap forward with their Emily text-to-speech voice, so much so that some callers will be unable to differentiate between a human and computer generated voice -- this really is progress," said Tony Tugulu, CEO, Tugulu, an information technology solutions company. "I am confident that my clients will see the immediate benefits of this exceptional new technology and I'm confident that take up by them will be high." SpeechWorks Solutions from ScanSoft offers a full suite of market-leading TTS solutions for embedded and network-based TTS applications that supports open industry standards such as SSML, MRCP and SAPI for empowering VoiceXML- and SALT-based applications, among others. www.scansoft.com/realspeak/ .

ScanSoft Announces Close of Rhetorical Acquisition: The acquisition of Rhetorical provides an array of technology, customer, partner and employee resources to help ScanSoft fuel its growth and propel the quality of speech applications throughout the world. In addition, Rhetorical Systems brings a range of benefits to RealSpeak Telecom 4.0 customers, including Active Prompts - optimized, expressive TTS output - supported by a suite of tuning tools enable customers to optimize application prompts throughout implementation and beyond, to ensure optimum accuracy. New release of French, US English, UK English, Polish and German languages to feature first names & last names lexicons by default; cities, streets and company names as options. www.scansoft.com

Telisma announces teliSpeech version of its Polish language, suitable for dynamic voiceXML applications: Telisma announced the availability of a teliSpeech version of its Polish language. This release includes a new, high-quality phonetizer and allows for easy development of sophisticated, dynamic services. An active promoter of open systems, telisma’s new product line, teliSpeech has been developed from the very start to be 100% standard compliant. The new Polish language has been created to take into account all the very strict VoiceXML constraints - something few other technologies really do. As a result, it is already available on all major VoiceXML Platforms, with no need for special integration or tuning processes. In Poland, telisma collaborated with PTK Centertel, helping the national mobile operator to launch the first voice message and e-mail consulting service ever deployed on such a scale in Europe in February 2004. The service provides users with friendly access to both their voice mail and e-mail via speech commands. It runs on a Unisys platform integrating telisma’s ASR and is available to hundreds of thousands of customers. www.telisma.com Contact: Pascale Garreau pgarreau@telisma.com +33(0_1 42 45 45 91

The W3C Multimodal Interaction Working Group Publishes Two New Working Drafts: The Working Group has just published a new Working Draft of the EMMA (Extensible MultiModal Annotation) specification. EMMA provides a standard, platform-independent means of representing user input and information about the input (annotations), such as confidence and timing information. EMMA is designed to accommodate user input derived from a variety of modalities, including speech, keystrokes and pen. In addition, EMMA can also represent composite input, which is derived from combinations of modalities. For example, a spoken input like "zoom in here", accompanied by a mouse click or pen gesture, can be represented in EMMA as a unified composite input. This specification can be found at http://www.w3.org/TR/emma/ . The Working Group has also recently published a Working Draft on a Dynamic Properties Framework. This document defines platform and language neutral interfaces that provide Web applications with access to a hierarchy of dynamic properties representing device capabilities, configurations, user preferences and environmental conditions. The device configuration, user preferences and environmental conditions can vary dynamically and applications need to be able to respond accordingly. For example, how should the application deal with low battery alerts or with loss of network connectivity? Perhaps the user has just muted the microphone or disabled the audio output? Dynamic configurations include snapping a camera attachment onto a cell phone or bringing devices together with Bluetooth, for example, a camera phone and a color printer. These dynamic changes to the configuration can directly affect the features that applications offer to users. This specification provides a framework for applications to access these environmental properties. This specification can be found at www.w3.org/TR/DPF/ . The Multimodal Interaction Working Group is very interested in comments and suggestions from the speech industry regarding both these specifications. Comments can be sent to the Multimodal Interaction Working Group's public mailing list, www-multimodal@w3.org .

HTI Voice Solutions, Inc. announced that Peter B. Keenan has been appointed as the Company's new President and CEO, succeeding Kenneth B. Brater, who will retire at the end of the year as the Company's Chief Executive Officer. CallMiner, Inc. announced that Mike Dwyer will head CallMiner software development. CallMiner also announced the appointment of Mike Pahner to the role of Region Sales Director. MedQuist Inc. (MEDQ.PK) announced that John W. Quaintance, the company’s Executive Vice President & Chief Operating Officer has decided to resign from the company.

*Supplier*	*Item*	*Q3, 2004*			*Q3, 2003*	*Q2, 2004*
*Supplier*	*Item*	$M	*2002 to 2003 % change*	*Q2, 2002 to Q3, 2003* *% change*	$M	$M
Artisoft	Total Revenue	2.90	45.0	11.5	2.00	2.60
	Net Income	(2.90)	29.3	262.5	(4.10)	(0.80)
	Cash or equiv.	7.26	34.0	164.1	5.42	2.75
Captaris	Total Revenue	18.00	(0.6)	(1.1)	18.10	18.19
	Net Income	0.07	-	(217.9)	5.90	(0.06)
	Cash or equiv.	88.10	1.0	50.2	87.24	58.67
Comverse	Total Revenue	245.48	26.6	5.2	193.84	233.40
	Net Income	9.54	(49.9)	28.3	6.37	13.30
	Cash or equiv.	2,199.50	4.5	0.6	2,103.94	2,186.76
Enghouse	Total Revenue	12.86	37.5	30.6	9.35	9.85
	Net Income	2.52	152.3	3.3	1.00	2.44
	Cash or equiv.	73.24	18.4	11.5	61.88	65.66
PROSODIE	Total Revenue	58.72			NA	NA
	Net Income	1.76			NA	NA
	Cash or equiv.				NA	NA
Transcend	Total Revenue	3.67	(2.3)	(2.5)	3.75	3.76
	Net Income	0.04	(86.1)	(58.1)	0.28	0.09
	Cash or equiv.	0.56	30.9	(0.2)	0.43	0.56
Verint	Total Revenue	63.99	30.6	6.4	49.01	60.17
	Net Income	5.18	10.9	(8.7)	4.67	5.67
	Cash or equiv.	225.60	0.9	(1.4)	223.69	228.71
Witness Systems	Total Revenue	35.30	26.5	2.6	27.90	34.42
	Net Income	2.94	-	94.4	(2.53)	1.51
	Cash or equiv.	62.95	61.0	20.3	39.10	52.35

PROSODIE publishes for the first time, as it had committed earlier, its non audited quarterly results for Q3 2004. As it is the first publication of quarterly results, there is no comparison except for revenues that where up 7.7% (8.5% at constant exchange rate and perimeter) for the quarter. The third quarter’s operating margin is always the highest of the year, in particular because PROSODIE Info has important volumes of traffic in the summer for the weather forecast and in July for the exam results.

*Company*	*Symbol*	*Share Price*			*2003*		*Market Cap. 12/1/04*		*Mkt Cap./Rev.*	*2002*
		*Last*	*52-wk.*		*Rev.*	*Net Income*	$	*% change*		*Rev.*
		*Last*	Hi	*Low*	$M		$	*% change*		$M
Andrea Electronics	AND	0.08	0.08	0.08	5.1	(4.6)	4.34 M	(38.44)	0.85	7.2
Apropos Technologies	APRS	3.41	6.61	2.57	20.1	(5.6)	59.71M	11.44	2.97	20.9
Artisoft	ASFTC	2.75	4.60	1.65	6.6	(7.1)	10.51M	0.00	1.59	4.5
Aspect Communications	ASPT	11.14	19.95	7.33	363.8	(29.0)	0.667B	(1.91)	1.83	396.1
Brooktrout Technology	BRKT	12.01	23.18	7.65	74.7	0.6	152.88M	(4.98)	2.05	73.5
Captaris	CAPA	5.16	6.70	4.04	83.3	12.5	156.69M	3.19	1.88	94.6
Cognitronics	CGN	4.09	4.50	3.00	10.3	(5.4)	23.42M	27.01	2.28	11.3
Comverse Technology	CMVT	24.45	25.07	15.25	765.9	(5.4)	4.83B	8.05	6.31	735.9
Convergys	CVG	14.99	19.96	12.30	2,288.8	171.6	2.12B	(0.93)	0.93	2,286.2
Covalent Group	CVGR	2.55	4.31	2.05	20.8	(0.6)	33.59M	12.34	1.61	29.2
Eckoh (UK)	ECK	8.00	17.70	6.00	104.8	(16.4)	40.2	(28.14)	0.38	104.8
Enghouse(Canada)	ESL	8.93	21.07	8.00	34.6	7.2	187.4	7.64	5.41	13.9
Fonix	FINX	0.193	0.66	0.12	2.4	(13.5)	21.94M	45.39	9.20	3.1
Glenayre	GEMS	2.18	4.30	1.43	58.2	1.6	146.82M	16.58	2.52	67.4
Interactive Intelligence	ININ	4.50	6.65	3.07	51.2	(5.9)	71.54M	(1.96)	1.40	47.8
Intervoice	INTV	13.35	18.00	7.23	156.2	(66.4)	480.83M	1.37	3.08	145.6
iVoice	IVOC	0.0005	0.0099	0.0002	0.4	(2.0)	3.3	(16.45)	7.36	0.6
NMS Communications	NMSS	6.31	9.35	3.71	87.1	(36.0)	298.99M	(4.39)	3.43	76.5
Nuance Communications	NUAN	4.14	9.07	3.53	55.0	(19.3)	147.74M	(5.69)	2.68	44.1
Preferred Voice	PFVI	0.145	0.33	0.07	2.1	(1.3)	4.29M	3.62	2.02	2.2
Prosodie (France)	PRD	19.39	28.21	14.80	203.5	(7.5)	156.2	(17.40)	0.77	203.5
SAFLINK	SFLK	2.78	4.76	1.98	2.0	(10.7)	220.5	(4.14)	111.38	2.0
ScanSoft	SSFT	4.19	6.36	3.25	135.4	(5.5)	442.69M	12.96	3.27	106.1
TALX	TALX	25.79	30.00	18.50	126.1	13.0	384.82M	(6.76)	3.05	137.8
VoxWare	VOXW	0.055	0.170	0.030	4.6	(6.6)	2.93M	(28.36)	0.64	4.6
Verint	VRNT	36.33	42.22	22.24	192.9	17.9	1.14B	(10.94)	5.91	131.2
Voiceware (ROK)		1.05	2.67	0.96	3.8	(1.6)	19.1	(13.42)	5.00	3.8
West Teleservices	WSTC	33.11	36.29	22.15	988.3	88.0	2.25B	(6.25)	2.28	820.2
Witness Systems	WITS	17.46	18.20	8.61	108.04	(20.6)	424.37M	5.84	3.93	67.69
Wizzard Software	WIZD	1.92	4.39	0.42	0.940	(2.0)	49.89M	27.99	53.07	0.346
Z-Tel	ZTEL	1.70	23.95	1.13	289.1	(16.1)	14.18M	(39.97)	0.05	235.3
Total	Total				6,246.4	23.3	14,565.9	0.37	2.33	5,877.9

The Emperor is Naked!
How well do we use it ourselves? How about regulating the VUI? Another approach to Name & Address More media bashing of speech systems Sensory’s new technology launch Contact Solutions doing well Bad-ideas die very slowly Another bad idea – Automatic LT	One-number Concept: Another bad idea A good idea – Speech improves customer service The Customer Made me do it!: Call Centers don’t know CC managers care about customer satisfaction? Persona Looniness Continues Speech Industry Resolutions for 2005 When outsourcing to Indians makes sense
Telephone IVR – Applications
Angel.com at adidas Contact Solutions/Intervoice Election Monitor Crealog/NMS in German Auction Eckoh/21^st Century at NIE Eckoh Wins Age Verification at O2 Fluency Voice at MM Group Metaphor Three new Customers Telisma at Jet Multimedia	Telus Outsources Maximus Health Services Vocantus/GEOCOMtms Delivery Scheduling Viecore at Palm Beach Schools Viecore at Medical Mutual of Ohio Viecore Insurance Claims Processing Voxify at Aer Lingus Swisscom Systems/VoxPilot at PostFinance
Telephone IVR – Products/Services
Apptera FS Packaged Apps Edify gets VoiceXML Certification Fluency Persona Selection	Paraxip New IP Telephony Products Vocalabs Improving customer service book
Telephone IVR – Partnerships/Financials
Acapela Eureka Project Diaphonics Provides VV to NAVA Telstra Outsources from CenterLink TuVox obtains funding	Unveil Partners with Genesys VOICE.TRUST/Telisma alliance Witness to acquire Blue Pumpkin
Telephone Messaging - Applications/Products/ Partnerships/Financial Partnerships/Financial
Avaya Ships Modular Messagig 2.0 Invores First Impressions IVoice at Bergen County Utilities Authority	HeyAnita RMS v2.0 Spanlink Provides IP system to law firm VoiceGenie/Neoris in Texas Schools
Telephone Operator Services - Applications/Products/ Partnerships/Financial Partnerships/Financial
Acapela in EDA Belgian Directory Service One Voice DA in Texas	Verizon 411 Struggling
Telephone Tools - Products/ Partnerships/Financial
Loquendo unveils Loquendo TTS Director ProNexus Upgrade for CT ADE	Paracon adds ProNexus CT and speech tools VoiceObjects awarded patent
Embedded Telematics – Applications/Products/Partnerships/Financial
Acapela joins Renasas ART selected by Gionee for phones NEC selects ART for HDM handset	Fonix VoiceDial in HTC Cellular devices Vodafone Expands ScanSoft SpeechPAK TALKS RNIB Distributes SpeechPAK TALKS
Embedded Appliances – Applications/Products/Partnerships/Financial
Japanese Robot for Elderly Sensory new speech technology	Wizzard Talking Pill Bottle Starter Kits
Desktop – Applications/Products/Partnerships/Financial
Readspeaker using Acapela Acepela in French Museum Inovani using Acapela in animation app Interactive Multimedia Vampire Lip Sync MacSpeech Transcription option Opera adds Voice to Browser	Right Seat Talking animated presentations Sakhr at Microsoft Summit in Cairo Vianeta adds ScanSoft DNS to Harmony Transcend adds ASR from MultiModal Wizzard Software License TTS to Asia Broadband
Data Entry – Products/Applications/Partnerships//Financials
AWG gets return from Vocollect
Base Technology – Products/Partnerships/Financial
Loquendo Dutch TTS Voices ScanSoft RealSpeak Telecom 4.0	ScanSoft closes Rhetorical acquisition Multi-modal working drafts from W3C
Industry People
Financials
Table F1 - Q3 Reported Financial Results	Table F2 – Market Price Summary-1/01/05
Calendar

*Item*	*Number*	*Greeting Length*	*Instruction Length*
Speech-enabled	57 (46.7%)	7.8 seconds	33.2 seconds
Touch-Tone	58 (47.5%)	8.2 seconds	64.9 seconds
Live	7 (5.7%)	2 seconds	4 seconds
Total	122 (100%)	7.4 seconds	45.3 seconds

Automatic Speech Recognition

ASRNews

Swisscom Systems/VoxPilot at PostFinance

*ASRNews***