Risk Manager

Acclaimed academic

When emcee Steve Jobs presented this year's Computerworld Object Application Award for Best Application Utilizing Reusable Components Leveraged From or For Use in Other Projects, Werner Staringer didn't realize the glass trophy was as hefty as its title. "The crystal piece was not fixed on the pedestal, so I dropped it on my way back. Almost everybody who gets it in hand for the first time does the same thing," the 30-year-old Ph.D. says.

Of course, Staringer's wasn't the only blooper at the ceremony, nor was he the only NEXTSTEP contender. Jobs needed a few seconds to pronounce the name of Staringer's sponsor, "Creditanstalt Bankverein of Vienna, Austria," and the list of finalists included afs:TRADE, an object kit for securities-trading applications from Springhouse, Pennsylvania-based Anderson Financial Systems.

Staringer's winning project was a financial Risk Management System (RMS), a program that helps Creditanstalt's managers monitor their exposure to fluctuations in rates of foreign exchange, interest, and inflation, or events, regulatory and otherwise. Rather than reusing objects from previous efforts, as most entrants in this category did, RMS stood out because it reuses entire applications Ð notably Lotus Improv for reporting and charting, Wolfram Research's Mathematica for its computational engine, and SQL (Structured Query Language) databases for its price quotes. RMS also won kudos for its flexibility and extensibility, since it allowed end users to build new kinds of securities and risk models on their own, using NeXT's InterfaceBuilder and Mathematica simulations.

Staringer is the driving force behind the Financial Markets Laboratory (FML) at the Technical University of Vienna, which built RMS in partnership with Creditanstalt. FML is one of 14 institutes within the school's giant Informatics department, but it's no cog in the academic machine. Like its founder, FML is an ambitious innovator. It is unique in its focus on real-world problems, its encouragement of student involvement (both graduate and undergraduate), and its funding, which is 95 percent corporate, rather than government, based. "We work on research, not products Ð although we are sometimes kind of in the middle. We are doing useful, practical things," he says.

When the consultants Creditanstalt had hired failed to deliver, the lab's team was eager to take over the project: "At that time, we were already working with NEXTSTEP, and we had experience with financial applications, and we wanted to do exactly what they needed done. They wanted NEXTSTEP, and they wanted people who knew this domain," Staringer says. The core group for RMS also included Otto Hainzl, who reverse-engineered the code that had already been written; Andreas Haleger, an expert in object-oriented design and GUIs; and Werner's brother Rainer, a UNIX wizard.

Werner conceived of FML when he returned from a yearlong teaching stint at the University of Auckland in 1987. Trained in expert systems, Staringer explains that "what always interested me about AI was not intelligence per se, but intelligent,'smart' programs," so he set out to explore an application domain rather than doing straight research. In fact, he had nurtured an interest in smart programs after his first encounter with serious programming, when he tried to teach an Apple to play Connect-5 using an early neural network.

Creating FML was a risky venture, but Staringer is no stranger to leaps of faith. He ran his own comic-book exchange as a youngster and later launched OKAY, a student newsmagazine that eventually grew to a nationwide circulation of 35,000. These days, he and his wife, the former CEO of the newsmagazine, have acquired a passion for small-plane flying.

Certainly, Staringer enjoys the acclaim in the business community from his recent award, but academics remains his first love. "I have run a company, and I know how much of a burden it can become. Maybe my views will change over time, but right now I value my academic freedom," he says.

by Rohit Khare