"Have you ever been to one of those huge industrial bakeries, like at Nabisco?" asks Carol Kovac, IBM's VP of life sciences. "It's more like a factory. They churn out 2 million boxes a day. I love that stuff."

The cookie-factory fascination stems from an early-'90s job running IBM's supply-chain software business, where Kovac gained an understanding of what it takes to mass-produce what scientists dream up. "There are lots of challenges to bringing a product to market," she says.

Kovac's facility with business and science gives her credibility with customers. Says Michael Keehan, chairman and CEO of NuTec Sciences Inc. in Atlanta: "With Carol, I can spend 90% of my time talking business."

Yesterday Kovac started her IBM career researching semiconductor circuitry in San Jose, Calif.; she moved to IBM's Watson Lab in upstate New York in the mid-'80s to work on interconnect technology for RISC machines Today Kovac helped negotiate IBM's joint venture with-and $10 million investment in-MDS Proteomics in May, to co-develop a protein database. "She takes you deep inside IBM," says MDS CIO Chris Hogue.

Kovac's unit is IBM's $100 million bet that the future's fastest computers won't always run big physics problems, but will crunch terabytes of data at companies that analyze the human genetic code and search for new drugs. Unlike at national laboratories, where most users are have IT training, biologists who buy from IBM and other vendors more closely resemble "pretty good hackers," says Chris Hogue, CIO at MDS Proteomics, a unit of Toronto health-care company MDS Inc., in which IBM invested $10 million this year. That means vendors need to take risks, along with their customers, by developing products rapidly in a sector where technology consensus is hard to build. "Carol grasps the idea that while standardization works for the IT industry, it's too slow for dealing with the exponentially growing amount of biological data," Hogue says.

Cutting through the product-development bureaucracy can be especially tricky at IBM. "Sometimes large companies put in place best practices to make sure mistakes don't happen again," says David McQueeney, IBM's VP of emerging business. "Carol is very good at respecting that but taking it with a slight grain of salt."

Her alliances within IBM also help. Kovac has forged friendships with vice chairman John Thompson, a protector of emerging businesses; Irving Wladawsy-Berger, who heads Linux and grid-computing efforts; and Mark Dean, an IBM Fellow working on an experimental project to build a supercomputer to analyze the complex problem of protein-folding.

All told, Kovac-who was a chemist at Union Carbide in the '70s before earning a doctorate in chemistry in 1981 and joining IBM in '83-is proving adept at navigating the corporate hallways. And that counts for users. "At the end of the day," says IBM's McQueeney, "you're involving customers in experimentation." Kovac's business gives them something to sink their teeth into as well.

Peter Thiel, CEO of PayPal Inc.	A World Of Opportunity
David Busser, CIO of the 2002 Winter Olympics	Winter Olympics CIO's Winning IT Strategy: Play It Safe
Michael Robertson, founder of MP3.com and Lindows.com	MP3.com Win In Hand, He Takes Aim At Windows
Richard Clarke, chairman of the President's Critical Infrastructure Protection Board	Tenacity Carries The Cybersecurity Message
Christopher Klaus & Thomas Noonan, Internet Security Systems Inc.	Transparency Is Apparent In Duo's Vision For IT Security
Michael Vatis, founding director of the National Infrastructure Protection Center	From Fear-Monger To Consensus-Builder
Dr. Paul Tang, chief medical information officer at the Palo Alto Medical Foundation	Patient-Privacy Issue Gets A Doctor's Care
Gene Tyndall, VP of global markets and E-commerce for Ryder	He's Fixed On Keeping The Supply Chain Strong

Photo of Kovac by Brian Finke