Microsoft Peppers the Market with Integrated Voice Solutions

Have you called your Web site lately? Microsoft yesterday made available its highly anticipated Microsoft Speech Server (MSS) platform, making way for multimodal (integrated speech, text, and Web) contact-center applications. The speech platform enables advanced speech-recognition solutions to retrieve text information from Web sites and relay it back to customers in a voice solution over the phone. Microsoft made the announcement with its usual fanfare. Before an audience of more than 4,000, Bill Gates introduced Microsoft Speech Server 2004 during his keynote address yesterday at the AVIOS SpeechTEK Spring 2004, Microsoft Mobile Developer Conference (DevCon) 2004, and Fawcette Technical Publications' VSLive! San Francisco 2004 events. MSS is based on the highly publicized speech application language tags (SALT) programming language, which enables companies to move beyond the technological restrictions of VoiceXML--a well-established open standards platform for developing integrated text and speech applications. SALT allows callers to listen to information obtained from a Web site. By introducing speech capabilities to Web developers, Web sites can be reachable through telephones, mobile phones, Pocket PCs, and Smartphones. "This is the most integrated single platform...for Web integration, speech, and text applications for the call center," says Xuedong Huang, general manager of product development for Microsoft's Speech Server Product Group. "We always look at improving the value [of applications] for our customers. The overall cost is not just the platform. It's the licensing of the platform, creation, and deployment [of the platform]. If you use VoiceXML, you have to rewrite the code for the Web." With the help of MSS, parents in the New York public school system could, for example, listen to their children's grades, lunch menu, and school attendance information over the phone, Huang says. However, is MSS ready for corporate America? "Enterprises committed to working in .NET [that] wish to develop in-house applications should consider MSS. Enterprises committed to other technology should track MSS, but should not feel compelled to change," says Bern Elliot, research area leader for contact centers at Gartner. "VoiceXML-based solutions have at least a two-year head start on [the advanced speech recognition] market. Speech recognition technology gets better as it processes more calls and the vendors continue to tune the acoustic and language models. VoiceXML solutions in 2004 will process over 10 billion calls in North America for large firms like AT&T, American Airlines, Federal Express, and etc. [MSS] has not had as broad exposure to such diversity yet," says Art Schoeller, senior analyst at Yankee Group. A key part of Microsoft's strategy is its partner ecosystem. Roughly 73,000 users have downloaded the Microsoft speech authoring tools, and about 1,000 MSS applications are currently being deployed, according to Huang. Some speech recognition vendors, including Intervoice, TuVox, and Voice Automation, have already come out with applications supporting Microsoft Speech Server 2004. With Voice Automation's Voice Automation for Microsoft CRM, for example, callers can update quotes, get contact information, listen and reply to emails, get appointments and phone numbers, and more. Already approximately 60 Microsoft partners have developed applications for MSS. "The most important thing here is Microsoft will invest a lot of marketing money to raise awareness about speech applications, which Yankee research confirms continues to be an issue in the market," Schoeller says.