A Free Spirit Provides a Firm Foundation for Growth.

In my last column, I promised to tell you about Yuval Davidor, the scholar and computer scientist whose algorithms solve intractable problems. Davidors company, Schema, is helping wireless carriers maximize the use of their networks, but thats just its latest incarnation.

Davidor founded Schema in 1994. Its focus on wireless—which began in 1998—is potentially lucrative, because while capital and spectra are scarce in telecommunications, the number of wireless users is growing. Although Schemas software optimizes CDMA networks, it is not limited to 3G or even wireless networks. Carriers like Verizon, a Schema customer, can save money while improving quality of service.

Schemas first big customer was the Israeli military, which optimized its missile systems. Davidor also figured out how the shipping industry could arrange boxes inside containerships—which resulted in a separate company—and sold technology to the medical industry that addresses the types of magnetic fields created during tissue scanning.

What is interesting about Davidor, aside from his brilliance, is his understanding of limits. Many people forgot this during the dot-com era, but limits are inherent in business, and respecting them can lead, ironically, to expanded possibilities.

Davidor asked himself what would make him wake up in the morning and enjoy every day. He wanted to do things that he felt were good. A potential investor asked him where he fit into an organization chart for Schema, and he had no answer. So he spent two years thinking about Schema—how it should run, how it should develop, what he should do.

Schemas technology can solve problems that introduce drastic change and affect large numbers of people, but there will be no more contracts with the military. "We will not do anything to inflict sorrow or misery to individuals or communities," Davidor says. Schema regards quality as a value in its own right, and works accordingly. And Davidor has structured Schema to run without him if necessary. As the company added new executives and directors, Davidor moved to Berkeley to immerse himself in Silicon Valley and think of new applications for his algorithms.

Schemas software costs more than that of competitors, but the company claims to deliver greater return, increasing network capacity by at least 20 percent using evolutionary algorithms that analyze a wide range of data. Schema compensates for flaws in the way customers operate—engineers do not usually include marketing or financial data in their network planning—and respects well-designed software as an entity in its own right. "We must know our roles and set softwares spirit free," he says. "We must know who is important and who is not."

Many things could go wrong with Schema, but Davidors spirit provides a firm foundation. This company will be fun to watch.