While many speech technology companies have thrown their support behind the SALT Forum, VoiceXML remains their bread-and-butter speech platform, and theyre closely watching efforts to integrate it with XHTML as an alternative to Speech Application Language Tags for multimodal application support. Heres a look at some of the established players in speech technology as well as up-and-coming startups focused on multimodal applications:
Nuance Communications Inc., of Menlo Park, Calif., develops what it calls VoiceWeb software—speech recognition, voice authentication, text to speech and voice browsing. The technology is frequently used for automated telephone customer service systems. Nuance also offers developer tools for building speech applications. Its VoiceWeb Server supports VoiceXML 1.0 and the developing 2.0.
Nuance backs the VoiceXML and Extensible HTML combination submitted by IBM, Motorola Inc. and Opera Software ASA to the World Wide Web Consortium. It doesnt support SALT.
SpeechWorks International Inc. is a founding member of the SALT Forum, although it also supports VoiceXML and has been very active in its development.
Based in Boston, the company develops natural language speech recognition, speaker verification and text-to-speech applications under the OpenSpeech brand. It works very closely with Microsoft Corp., generally acknowledged as the driving force behind SALT, to build applications on Microsofts .Net Speech Platform.
VoiceGenie Technologies Inc. makes VoiceGenie VoiceXML Gateway, a Unix-based platform for running speech applications, including telephony interface cards, automatic speech recognition and text-to-speech applications.
VoiceGenie, of Toronto, also develops a voice browser and other applications such as contact management, voice- activated dialing, e-mail by phone, voice mail management, a calendar and reminders. Though a VoiceXML supporter, VoiceGenie has also joined the SALT Forum.
Voxeo Corp. develops natural language voice applications for customer self-service. Its Voxeo Voice Center platform supports speech recognition, text to speech, VoiceXML processing, call center integration and call control features.
Though it supports VoiceXML, Voxeo, of Scotts Valley, Calif., has also joined the SALT Forum.
Telera Inc. develops VoiceWeb Application Platform, designed to link the Web with telephony systems. It lets users access Web-based applications via voice from a telephone. These applications can include Touch-Tone and speech-activated Interactive Voice Response, voice recording, queuing, routing, and outbound notifications.
All of these applications are developed and deployed using open standards such as VoiceXML. In fact, Telera supports the VoiceXML 2.0 specification now being developed by the W3C Voice Browser Working Group, of which it is a member. But like other VoiceXML supporters, the Campbell, Calif., company has also joined the SALT Forum.
Kirusa Inc. and Lobby7 Inc. are startups focused on multimodal applications.
Kirusa, based in Berkeley Heights, N.J., is building the infrastructure to support wireless multimodal applications for carriers and enterprises. It recently launched the Kirusa Application Development Program to provide technical and marketing support, including a development and testing environment, to third-party developers, partners and customers using its MultiModal Platform to build wireless multimodal applications. The company has joined the SALT Forum, though it also supports VoiceXML and recently announced partnerships with Telelogue Inc. and VoiceGenie to add multimodal capabilities to those companies voice technologies.
Lobby7 develops application servers that handle multimodal inputs and outputs between users and devices. These products include the Xmode Network Server, for handling such interactions between devices and appliances that use a network connection, and Xmode Embedded Server, which runs on the device or appliance used and enables a multimodal interface without a live network connection.
Lobby7, of Boston, has joined the SALT Forum, but its products also support VoiceXML and XHTML.