|
|
Nervana ~ News ~ Articles
Software to help busy people beat data overload
Tricia Duryee | Monday, January 10, 2004
Who: Nosa Omoigui
What: Nervana, a six-person company developing information-management software in Bellevue.
Web site: www.nervana.com
Background: Omoigui, who was born in Nigeria, spent seven years at Microsoft, where as a program manager in Microsoft Research he filled several roles, including developing ideas for .NET. After filing for 14 patents in one year, he said he was burned out, which prompted him to leave in 2001 to start his own company.
Birth of an idea: While at Microsoft, he received more than 3,000 e-mails a day. Omoigui said the load was unmanageable, but if he took his name off any of the department lists he was on, he was left out of the loop. "Your inbox has no smarts whatsoever. It's like a fire hose being pointed at you and hitting you in the face with no screen," he said.
Doing something about it: Based on his experiences, Omoigui started Nervana in 2001, filing, yes, another patent (in all, he has been granted two and has 20 pending). This 300-page patent application was titled the "Information Nervous System," a design to reduce the pressure from a blast down to a trickle.
Google's drawback: Nervana is similar to Google in that it finds information, but Google isn't as intelligent. It relies on users to express what they want in a search box, Omoigui said. For example, when you type in SARS, you get 6 million results, and the fourth or fifth choice has nothing to do with the disease but is about the South African Revenue Service. That's fine when you are searching for a used guitar, the 32-year-old Omoigui said, but for a knowledge worker who's sifting through thousands of documents, research papers and e-mails at a large company — as he had to do at Microsoft — it's highly inefficient.
Unlike Google: Nervana is designed to search by relevance and meaning rather than by keywords.
Word lesson: The technology is powered by taxonomies and ontologies. A taxonomy classifies words in groups by their relationship. Ontologies group words and concepts by meaning.
How it works: Users create profiles detailing who they are and what kind of information they want and from what sources. They then click on folders such as "All Bets," "Best Bets" and "Breaking News" to find relevant, timely information.
Competition: There is a movement called The Semantic Web by the World Wide Web Consortium, or W3C. The Semantic Web supports the idea of searching by meaning and not keywords. However, the drawback to many of these solutions is that they require documents to be tagged and coded so they can be easily found. Omoigui said that's impractical; there's no way to go back and tag everything in existence or create standards going forward.
Funding: It took him two years to raise $800,000. Investors include former top Microsoft executive Scott Oki and WestRiver Capital. Omoigui plans to close on a $1.2 million financing round soon to launch the product and sales effort.
|
|
|
|