Q&A: ISVs Take Microsoft Speech Server 2007 for Test Drive

REDMOND, Wash. – May 8, 2006 – Microsoft today announced availability of the beta version of Microsoft Speech Server 2007, which is scheduled to be released late this year. Beginning today, beta testers will have the opportunity to preview and provide comment on the release. Microsoft Speech Server 2007 is the company’s next-generation speech and telephony platform. Its purpose is to help companies achieve two goals that might seem to be at odds: reduce costs and improve customer service.

To learn more about the product, and call-center technology in general, PressPass spoke with three independent software vendors (ISVs) who specialize in voice-recognition solutions: Andrew Watson, consultancy director, and Jamie Prady, senior application architect with Quorum Network Resources, Ltd.; Joseph Carey, president of Versay Solutions, a subsidiary of Vail Systems; and Klaus Zuenkler, vendor and technology manager, TietoEnator, Digital Innovations.

PressPass: What’s your company’s specialty?

Andrew Watson (Quorum Network Resources): Our specialty is advanced infrastructure. We provide consultancy to financial institutions on Web-server hardening and Internet security, which comprises about 70 percent of our business. The remaining 30 percent is the development of applications, which we write for the finance industry. We were founded in 2000 by Charles Scott and Andrew Watson. We have 15 employees, are privately owned, and based in Edinburgh, Scotland, which is where the majority of our clients are.

Joseph Carey: Vail Systems is an application services provider (ASP) that hosts both directed-dialog and natural-language speech applications, as well as VoIP SIP-based telephony services such as conferencing and intelligent call routing. In partnership with our subsidiary, Versay Solutions, we also provide a full suite of professional services related to voice application development, voice user interface design, persona development, computer-telephony integration, and Web-services data integration. We were founded in 1991, and I’m proud to say that 15 years later we still have our very first customer. And don’t let the name of our company mislead you: we’re located in Chicago, not in the ski resort in Colorado with the same name.

Klaus Zuenkler (Digital Innovations): We are one of the largest IT services companies in Europe. We offer a leading-edge know how that is geared toward developing innovative IT solutions that realize and digitalize the visions of our customers. We work in close partnership helping them manage and run their businesses more effectively. We’ve chosen to focus on areas where we have the deepest expertise, including global banking, telecom, healthcare, forestry and energy. In those areas we are working hand-in-hand with many of the world’s leading organizations. We are growing with them and are now active in more than 25 countries. Our long-term experience in telecom and data center infrastructure consulting, deployment and operations in combination with our future office and mobility solutions will help customers achieve everywhere access to e-mail, voice mail, calendar and contact data via a telephone, mobile device or a PC to increase overall productivity.

PressPass: What are some of the ways that customers, end users and callers suffer when they’re faced with poorly written applications?

Zuenkler: There are still a lot of poorly designed applications with a lot of common design mistakes. One of the more common is the very simple problem of users forgetting the correct digit to press if the announcement is composed in the wrong order. For example, “please press 4 to speak to customer support for printers” instead of first announcing the service and then the action to be taken. At the more difficult end of the spectrum is the fact that handling technical, dialog and recognition errors requires very sophisticated handling strategies that are often underestimated and not implemented in many systems. That leads to customers trying to avoid calling systems where they get stuck in the famous “voice mail jail.”

Watson: Poorly written applications can badly misrepresent a company. The idea of a speech or Web-based application is to automate a common area to make things slicker and more “self-service” in nature for customers. A badly written application that leads the user around in circles can cause a great deal of frustration and lead customers to believe that your service is suffering due to cost cutting. Basically, a well-written application will empower your users, while a badly written one will make them feel helpless.

Carey: For customers, we find that time to service is the biggest pain point. A poorly designed application frustrates callers and often causes them to be transferred to an operator either because of too many errors or by choice. I have to say, we’ve all been guilty of punching “zero” when faced with an unnecessarily complicated application or because of too many errors. Ultimately, though, that defeats the promise of self-service speech technology. We’ve also found that well designed applications perform a specific task and then send customers on their way. Our most satisfied customers remember completing their request successfully – not the technology that enabled them to do so.

PressPass: As an ISV, what attracted you to Microsoft Speech Server 2007?

Watson: It’s the ease of development, the low cost of ownership, and the low maintenance for our customers. There’s an opportunity to develop our speech business into areas where, previously, cost was a barrier to entry. Also, we developed and delivered some of the pre-beta courseware for the first version of Microsoft Speech Server and, in conjunction with Content Master, written the official training course for Microsoft Speech Server 2007.

Zuenkler: Microsoft Speech Server 2007 fits very well into our overall corporate strategy of building a more efficient information society, and our special services and product portfolio of digital customer interaction systems. Also, we agree with Microsoft that Microsoft Speech Server 2007 will conquer the market due to its capability as a very easy system to integrate, its dialogue design, and the ease with which it can be deployed. We believe that Microsoft has taken a really big step forward with Microsoft Speech Server 2007 when compared with current systems on the market.

Carey: As a long-standing Microsoft partner – since the first preview of Microsoft Speech Server 2004 – we are very familiar with the product, which naturally made us want to be part of the Microsoft Speech Server 2007 TAP program. We find the value proposition of the Microsoft Speech Server platform – both as an integrated stack and in relation to application development tools available with Microsoft Visual Studio 2005 – aligns well with our application development lifecycle framework and related product offerings.

PressPass: What are you planning to develop on Microsoft Speech Server 2007?

Zuenkler: In keeping with our focus on contact centers, we will be developing applications for voice portals for customer information and transaction access as well as automated attendants for larger organizations.

Carey: We’ve been working very closely with Microsoft on a number of fronts. We are porting several of our VoiceXML applications and also evaluating Microsoft Speech Server 2007’s built-in conversational understanding technology for several natural language speech projects currently underway. We are also beta-testing the Vail CTI Gateway, which will enable enterprises and application developers to connect Microsoft Speech Server 2007 to industry-leading CTI systems.

Watson: We are developing a repair tracking application for the mobile telephone industry and have some other development prospects that will firmly establish this sever in some key places in the financial industry. We are also organizing a demo in conjunction with Microsoft Scotland for a major regional bank.

PressPass: How will what you’re developing help your customers?

Zuenkler: The main advantage for our customers is that they will be able to keep pace with new voice-portal services with a faster time to market along with lower maintenance costs.

Watson: The applications we’re developing will reduce the amount of operator time spent dealing with tracking repair status and allow operators to focus on higher value and more interactive calls. We think we can take their overall cost per call down to 30 percent of the current cost with a minimal hardware and software expenditure.

Carey: The solutions we have developed for Microsoft Speech Server 2007 use technology ranging from VoIP and CTI to VoiceXML, SALT, and Conversational Understanding, which helps Microsoft Speech Server customers, partners and developers take advantage of all the great benefits of Microsoft’s platform while also giving them a greater degree of flexibility in terms of where and how they want to use it. This flexibility is essential in today’s call center environments, where legacy systems co-exist with next-generation VoIP networks and multi-vendor systems. For example, the Vail CTI Gateway solution helps enterprises offer better integration between the automated self service, and the agents, which in turn leads to increased customer satisfaction, shorter call lengths, more efficient call routing, and improved analytics. At the end of the day, these systems need to work in harmony and provide the caller with the best experience possible, while at the same time making sure we keep an eye on costs.

PressPass: What other Microsoft technologies are you using, and how do they integrate with Microsoft Speech Server 2007?

Zuenkler: We are using a lot of core technology from Microsoft, including, but not limited to, the .NET platform and Microsoft SQL Server 2005, all of which are integrating into Microsoft Speech Server 2007 very easily.

Carey: We use Microsoft SQL Server 2000 and 2005, Live Communication Server and Microsoft Office Communicator, IIS Web Servers and Active Directory. We have integrated with Microsoft SQL Server and Active Directory for unified messaging and communications, integrated with LCS for some breakthrough, and with novel VoIP applications, and in many scenarios we’ve integrated with .NET Web servers and Web services as major components of open-standards, back-end integration. We hope to continue this level of diverse and seamless integration with the technologies that will be delivered as a part of Microsoft Speech Server 2007.

Watson: We are pretty much a full-on Microsoft house, and most of our development work is done in Microsoft Visual Studio. The integration between Microsoft Visual Studio and the speech development functionality in Microsoft Speech Server 2007 makes for a very valuable proposition, especially when combined with the ability to integrate with client’s existing C# code base and business logic. And add in Microsoft SQL Server 2005 and you’ve got everything you need to develop a fully integrated and dynamic speech application. In much the same way we’ve done things for the Web, we can now plug speech in there quickly and easily.

Related Posts