Sophisticated computer systems are revolutionizing how institutional investors trade stocks -- why not the way they decide what to buy and sell? That's what Glen Wolyner, a former Bloomberg and Nasdaq technology executive, is betting as he seeks to breathe new life into former dot-bomb Revere Data. Launched as Sectorbase.com by former hedge fund manager David Baker in 1999, the San Francisco company quickly burned through its venture capital financing while trying to compete with giants Reuters and Bloomberg in the business of offering quote terminals. Since signing on as CEO in November, Wolyner has slashed overhead and focused on expanding Revere's research clientele.

The firm's main attraction is a proprietary database that classifies thousands of public companies by industry sectors, customers, suppliers and even specific products. Users can generate trading ideas by querying the system on these criteria. An investor, for example, might screen for all the known suppliers of a company that just announced a dramatic decline in revenue and then sell those stocks short. Revere employs 25 analysts who constantly scan regulatory filings, news releases and other public documents to keep the database current. A basic subscription costs $750 a month.

"Hedge funds have really taken to it," says Wolyner, who helped Revere end 2004 with its first-ever break-even quarter and projects that it will turn a profit early this year. About half of the company's more than 200 research clients are hedge funds, he notes. The rest include proprietary and program trading desks as well as other independent research firms. Last month Wolyner hired sales teams in New York and San Francisco that will work to sign up mutual funds and other traditional asset managers. Other prospects include mergers and acquisitions bankers and internal corporate M&A teams, who can use the system to identify potential takeover targets.

"We believe this is where research is going in the future," says Wolyner. "We're never going to tell you to buy, sell or hold. But we're going to draw you to clean data that will allow you to come up with good ideas."