U.S. Firm Finds Top Soviet Talent--At Bargain Prices

WASHINGTON -- A Boston University computer scientist and a Baltimore software company have put an unusual twist on United States firms' practice of going abroad in search of cheap labor. For Yuri Shestov and Intelligent Resources International (IRI) Inc., the underdeveloped world is the Soviet Union. "A lot of companies don't realize how much scientific talent is available in Russia," says Shestov, vice president of research and development at IRI. "And it's unbelievable what you can pay for labor there, compared to the West. Shestov, a U.S. citizen who emigrated from the Soviet Union in 1974, is well aware of the enormous reservoir of talent that exists in his native country. Once a mathematics teacher at Leningrad University, the 42-year-old Shestov has stayed in touch with many of his former colleagues since coming to the U.S. In the past few years he has been able to travel more frequently to the Soviet Union, and it was on one such trip to Leningrad last fall to see friends that he discovered the extent of this untapped scientific resource. "I happened to mention that my company was looking for some help and wondered if they might be interested," says Shestov, whose work involves modifying traditional computer software systems to take advantage of the speed and flexibility of neural networks. "The response surprised me. They were quite eager to do whatever I suggested. In less than a year, IRI has put some of his colleagues to work on six projects. A large team of several dozen programmers and scientists from a technical institute in Leningrad is tackling one project as part of a joint venture set up between IRI and a Moscow-based organization. And Shestov has organized smaller groups of scientists in each of the Baltic republics -- Lithuania, Latvia, and Estonia -- who are eager to demonstrate their skills. Each team includes prominent members of that country's mathematics and computer science communities, he notes. The six projects have a total budget of about $250,000, Shestov estimates, although he says that the work being done would cost IRI more than $1 million if purchased in the West. Shestov says that Soviet scientists are willing to work at wages four to five times lower than U.S. scientists because the pay is still three or four times higher than what they would receive for comparable work in the Soviet Union. "This kind of expertise costs $60 an hour," he says, "and in the USSR they would get paid the equivalent of $3." But Shestov is not pursuing collaborative ventures simply for the money. He says that there are equally strong humanitarian reasons for having U.S. firms look East for top-notch scientific talent. "In Vilnius, for example, scientists are eager to sign contracts with us just to get bread on the table," Shestov says. "The universities there pay their faculty very little, so it's nice to give them the opportunity to earn a little extra money at the same time they can exercise their skills." IRI officials give Shestov full credit for making possible such a collaboration. "If it weren't for Yuri, we wouldn't be involved at all," says Julia Stalfort, director of business development for the two-year-old company, which has a work force of 35 and $4 million in annual sales. The company was founded by entrepreneur George Silvais, who had also created another company, Terminals Unlimited, with which Shestov had previously worked as a consultant. Asked what U.S. businesses should know before trying to link up with Soviet scientists, Shestov doesn't mince words. "It would be much harder for anyone without a knowledge of Russia to arrange such a collaboration," he says. But he does offer some practical advice. "The biggest mistake that people make is trying to start at too high a level," he says. "Don't go to the state ministry -- you'll never get anything done because it's too bureaucratic at that level. Instead, you need to go to the level of a working group within an academy or an institute. "Go to the people who have the actual responsibility to make decisions and control budgets -- the chairman of a department of computer sciences, for example," he suggests. "Of course, sometimes they may need to verify it with someone higher up. But they know how the system works, and how to deal with any problems. You don't. What you want to avoid is meeting hundreds of people, only one or two of whom is in a position to help you get something done." Shestov says that IRI and Soviet scientists communicate regularly via electronic mail, as well as by fax and telex. The Soviets are paid in rubles, says IRI's Stalfort, although they also receive such in-kind services as computer equipment and training. Shestov hopes that the IRI collaboration will trigger other efforts to strengthen ties between Soviet and U.S. computer scientists. For example, he is seeking to have Soviet scientists contribute programs to the Cambridge, Mass.-based Free Software Foundation, which is coordinating volunteer efforts to create a new, public-access operating system that is compatible with the increasingly popular Unix operating system. FSF founder Richard Stallman, an independent computer scientist, says that he has agreed to accompany Shestov on a trip to the Soviet Union this fall. "I'm always happy to explain what we're doing," says Stallman, who last month received a five-year, $240,000 "genius" award from the John D. and Catherine T. MacArthur Foundation for his efforts to entice scientists worldwide to contribute free software. But Stallman believes that the era of cheap Soviet labor may be a short one if scientists in the USSR begin to adopt "capitalist" ideas about the value of their expertise. It could also bode ill for his own project. "If everybody has become property-mad, like you hear, then they may not be interested in writing shared software," he says. "But if they're really interested in writing good software that relates to the new system, then I'll be happy to talk to them."

U.S. Firm Finds Top Soviet Talent--At Bargain Prices

Interested in reading more?

Become a Member of