Ignored by Business, Data Manager Prometheus Goes Back to School

In 2009, Leon Rozenblit was ready to expand Prometheus Research from building databases for biomedical research departments at such universities as Harvard and Yale to energy and financial service businesses. Within a year and a half, he had burned through $1.5 million and laid off the five employees he had hired for the expansion, unable to penetrate Corporate America’s bureaucracies. “It was an expensive education,” he says.

Now he’s taking his 40-employee New Haven company back to its roots in the ivy towers of academia, selling software for scientists to use to organize and share data gathered in research projects. While unable to invest as much as multinationals, universities and foundations that fund academic research are increasing their use of technology that enables collaboration over the Internet. It’s one sign of a cultural shift within institutions known more for competing than sharing. “This is an entirely new way of thinking,” says Stephen Johnson, a professor of biomedical informatics at Columbia University. “People are now thinking about science as a collective enterprise, rather than a single scientist doing a single experiment.”

Rozenblit is confident Prometheus can win business with technology that’s flexible enough to adapt to the unique types of data gathered by biomedical researchers, such as genetic information drawn from blood samples. “Our bread and butter is supporting these very complex -- and I don’t want to sound derogatory -- messy academic projects,” he says. Still recovering from its unsuccessful foray outside academia, Rozenblit says Prometheus had $3.6 million in revenue in 2010 and expects $3.8 million this year. Starting in 2013, he’s betting the shift in university and foundation computer spending will drive annual growth of 30 percent to 50 percent over the subsequent few years.

Of course, the earliest stages of biomedical research -- the type supported by Prometheus -- gets the least amount of financial backing because it often advances the knowledge of a disease without producing a marketable treatment, says Alan S. Louie, analyst at market researcher IDC Health Insights in Framingham, Mass. While the technology market for academic research is small -- most likely in the tens of millions of dollars today -- Louie believes it will grow significantly over the next five years.


The Simons Foundation in New York is adamant about the benefits of having researchers across disciplines work together. Over the past five years, the nonprofit has spent $12 million on software and computers for an autism research project that involves 13 universities gathering and sharing genetic information from 3,000 families. The purpose is eventually to identify all the variants within genes that can be linked to the disease, explains Gerald Fischbach, the initiative’s scientific director.

The foundation hired Prometheus in 2006 to build the central database for the project after its founder Jim Simons (also the founder of hedge fund Renaissance Technologies) attended a company presentation at Yale. “We felt Prometheus had a good understanding of how scientists think and work,” says Johnson, who besides teaching at Columbia was hired by the foundation to manage the building of the computer system.

Prometheus’ proprietary software, called HTSQL, made it possible for Simons’ researchers to search and retrieve complex sets of data over the Internet using only a Web browser around the time the idea of cloud computing was becoming widespread. “There are things somewhat like it, but at the time for this space in which we were working -- scientific data management -- it was unique,” Johnson says.

Rozenblit founded Prometheus in 2000 while working on a doctorate in psychology at Yale University. “I was the nerdy grad student that everybody else asked to help with their data problems,” he says. Rather than work for free, Rozenbilt started charging the university for helping researchers configure their databases to match the specifics of their projects. When he earned his PhD in 2003, he set to work turning Prometheus into a full-time business, targeting other universities.

While Prometheus failed in the corporate market two years ago, Rozenblit believes HTSQL still has a chance of becoming a “transformational” technology for businesses. This time, instead of hiring a team to bang on doors, he’s partnering with other tech companies in data management. In early May, Prometheus announced it would work with nine-person database application developer Cirrus Technologies in Brick, N.J., combining the companies’ software to make so-called business intelligence applications for analyzing data. “If the technology gets accepted in the market on a large scale, it’ll happen over the next five years,” Rozenblit says. “If not, it will become a niche technology.”

To contact the reporter on this story: Antone Gonsalves at antonegonsalves@gmail.com

To contact the editor responsible for this story: Nick Leiber at nleiber@bloomberg.net

Press spacebar to pause and continue. Press esc to stop.

Bloomberg reserves the right to remove comments but is under no obligation to do so, or to explain individual moderation decisions.

Please enable JavaScript to view the comments powered by Disqus.