It's a familiar scenario: Professor and grad student hit on promising technology, decide to test the free enterprise waters, and emerge successful business executives. But in the case of Eric Brewer and Paul Gauthier, this textbook situation understates the dimensions of the payoff.
Propelled by the dynamics of the hottest sector of a buoyant economy, Inktomi Corp. has grown from a 10-person outfit in 1996 to a more than 800-person Web-technology player with 23 offices worldwide.
Underlying this tech-business narrative is bedrock Internet architecture technology -- parallel systems and search engines and a horizontal integration strategy. Brewer and Gauthier's investigation of scalable Internet server architectures led to the creation of a search engine and site that were to become the basis of Inktomi.
"We had an opportunity to not only affect the search engine space but the architecture of the Internet," Brewer explains.
Inktomi focused on the search engine space, looking "to be the engine behind the search engines," Brewer says. The company then broadened into related Web infrastructure areas, eschewing typical Web branding and product strategies.
"Pretty much at the time, every new Internet company was in some sense vertically integrated. It was a site and a technology producer, a customer service center. I think it's fair to say ours turned out to be the better approach," he says.
In 1997, the company branched out into network performance products. "That was a pretty big leap. We were less than a year old, and we decided to split the company," he says.
Reflecting his current academic focus on Web architectures, Brewer notes the key role of so-called edge-of-network servers.
"It's a new layer that in some sense does routing, but does a lot more other things, such as management of content, log collation, profiling, and distribution," he says. "It's a more powerful and more complex layer, but I think the Internet is ready for it."
Brewer is also focused on larger issues facing the high-tech sector.
"There's privacy, then there's copyright rights," he notes. "Technology [alone] is not going to provide a solution. It can help with some things like tracking and auditing and fingerprinting -- which kind of marks the content so that it can be proved to be of a particular source. But one key element is social, and certainly that's quite ongoing."
Overall, Brewer's grounding in the small-scale, intellectually engaging settings of lab and seminar, and their Web-economy parallel, the start-up, has made the adjustment to Inktomi's growth somewhat difficult.
"Some of the ways we operated at 100 [employees] are no longer appropriate. We try to keep people in relatively small groups, where they can feel part of something that's small and fast-moving and agile. But it's an ongoing challenge," he says.
Current position: Co-founder and chief scientist, Inktomi; associate professor of computer science, University of California at BerkeleyTechnology predictions: Edge-of-network servers will predominate, copyright issues will persist, and horizontally integrated technology business model will prevail