Cognitive Code: Speech recognition goes mobile

Service: Cognitive Code

What it does: An artificial intelligence platform for conversational speech recognition. While it's not a Web service, it can be applied to networked technologies, and has a lot of potential in the mobile world.

Why it will succeed: Speech recognition software has not taken off for consumer applications, largely because software has to undergo a lengthy training process keyed to individual speakers. Moreover, most programs require a lot of processing power, storage and memory. Cognitive Code's SILVIA (symbolically isolated, linguistically variable intelligence algorithm) software appears to be well on its way to overcoming those limitations, according to its creators, since it's speaker-independent and runs on the kind of low-power processor you'd find in a handheld. SILVIA is also one of the first (if not the first) software of its kind to actually "understand" words in context. Although the company initially released a desktop platform, showing it off during an impressive demo at last year's TechCrunch40 conference, the company's future is likely to be tied to small devices, including mobile phones, PDAs, toys and embedded gadgets.

Why it hasn’t succeeded yet: Cognitive Code's technology is new, and the first mobile OEM products aren't expected until late 2008.

'Net views:

CrunchGear: "The impressive element of this technology lies in its embeddability. Imagine, for instance, that SILVIA was embedded in a robot with sight and movement capabilities. Ordering it to "go get a red block" would send it on a search for a red block and it would then return to you with a block. Theoretically, anyway."

Management: CEO Elliot Polanco, Leslie Spring, CTO, Mimi Chen, CIO. Privately funded.

Related prediction:

• Prediction: Cognitive Code's Pocket SILVIA technology will appear in mass-market consumer electronics by end of 2008

Other related content:

• Slideshow home: Opinion: 10 'Net services that will succeed (and 10 that will probably fail)

Note: Anonymous comments on The Industry Standard are disabled. To leave a comment and participate in the Standard's prediction market, please register first.