Microsoft aims to add speech-enabling technology to a future version of its Exchange Server as part of its unified messaging strategy, a move that could potentially compete with its third-party ISV (independent software vendor) partners.
Microsoft unveiled plans to integrate technology from its Microsoft Speech Server into Exchange at the SpeechTEK Exposition and Educational Conference in New York on Tuesday.
The addition of speech technology to the software giant's messaging and collaboration server would enable a host of next-generation unified messaging scenarios, such as business users listening to their e-mail being read to them by an automated system over a mobile phone, said a company spokeswoman.
Further details on how the integration will take place have yet to be ironed out, she said. The next version of Exchange, code-named Exchange 12, is expected to be available in the second half of 2006, but Microsoft has not decided whether speech technology will be part of that release or the one following it, the spokeswoman added.
At its most basic level, unified messaging means e-mail and voice-mail messages are stored in a single mailbox, and can be accessed through whatever client -- mobile phone, laptop, personal digital assistant -- is available to a user at any given time.
Exchange is Microsoft's platform of choice to provide unified messaging, said Rob Helm, research director with Directions on Microsoft in Kirkland, Washington. But one problem unified messaging poses is that the number of messages a user can access in one place could be too large to slog through in a timely fashion. Some filtering of those messages according to relevance may be needed, he said.
"Microsoft is trying to position Exchange as the big inbox in the sky for everything," Helm said. "But the question is, who's going to read all that stuff? You may need automated systems."
This is where speech enablement can help, he said. Until now, speech technology has worked best on specific tasks, such as speech-enabled company name directories and automated customer service systems, that have a limited vocabulary. But extending e-mail systems with speech enablement could provide new scenarios for providing unified messaging, Helm said.
"Speech has worked best in cases where it’s a well-defined task," he said. "I expect the real value may be coming from having an automated system read your e-mail for you and route it to the right place, or read your voicemail and decide what to do with it."