New York (NY) - At an early morning session of Day 2 of the SpeechTEK conference, which is focused on speech recognition technologies, Microsoft demonstrated more of the integrated speech recognition features that the company plans to unveil in its upcoming Office Communications Server 2007, an enterprise networking extension to Office. After this component is released in the second quarter, users of corporate networks may find themselves communicating with Outlook and Exchange not only through their local microphones, but from their mobile handsets as well, using voice communication.
Exactly how much of the system was demonstrated is still unclear, even from the wealth of material Microsoft made available to the press outside the conference enclave this morning. But if the demonstration by Richard Bray, the company's general manager for its Unified Communications platform, had any more success than presenters at TechEd 2006 two months ago, attendees probably saw a beta of Outlook 2007 presenting a voice mail message to a user. If Bray was luckier still, he managed to pull off the part of the demonstration where Exchange communicated with the voice mail sender via telephone, using an automated voice coupled with speech server technology that recognizes the sender's responses.
The demo came by way of announcing that Microsoft plans to integrate its Speech Server 2007 capabilities into Office Communications Server. This was fully anticipated, mainly because Speech Server doesn't appear to be designed to work with much else, but also because it doesn't make too much sense to have "Unified Communications" without the products that deliver it being unified as well.
In one of Microsoft's corporate Q&A statements released this morning, Anoop Gupta, the corporate vice president in charge of the Unified Communications Group, said as much: "The speech and telephony capabilities provided by Speech Server create an even more robust platform for customers to build from - essentially they will be able to manage all of their communication needs from one platform, helping to save money on tools, training and the overall headache of having to manage multiple systems."
Gupta once again presented the metaphor of digital communications being centered around multiple "silos," each of which is geared to its unique mechanism - office e-mail, wireless e-mail, voice, VPN conferencing. The model Microsoft is trying to extend compounds the silos into a single platform, which is essentially a context-neutral carrier of digital communications. "Adding Speech Server to Office Communications Server will give our customers the tools to provide seamless access to applications and information from any device, anytime, anywhere. Outlook Voice Access and Unified Messaging in Exchange 2007 serve as powerful examples where integrated speech capabilities will expand access and unify email and voice mail."
Speech recognition becomes critical in this model, Microsoft argues, because users who do not happen to have direct access to computer keyboards still need control over the communications process. As a case in point, a traveler on a conference call may still require the ability to order up an e-mail of a recording of the call to another invitee who was unable to attend. If Outlook or Exchange could recognize when you're talking to "it" ("him?" "her?") rather than to someone else on the call, it could accept the traveler's command verbally rather than through a graphical front-end.
Today's demonstration does appear to have been a scheduled part of the regular conference program, as opposed to the case yesterday, where a "presentation" Microsoft made to a search engine technology convention, according to the photo blogs of those attending the conference, did not appear to have taken place in person.