by Brian Caulfield

SynerScope will join several dozen other startups sharing their ideas at the Emerging Companies Summit, at the fourth annual GPU Technology Conference, to be held in San Jose, Calif., next March.

Jan-Kees Buenen pulls an iPhone out of his pocket. “Everyone talks about networks,” says Buenen, a Dutchman with an unruly mop of curly brown hair. And thanks to smartphones, we all carry around these networks with us in the form of contact lists and call logs… not to mention Facebook and LinkedIn. “But who has seen his own network? Who even has an idea of what his network looks like?”

Big data is a powerful trend. But it’s usually measured in eye-glazing abstractions. Words fail to capture how the ability to crunch petabytes of data is revolutionizing fields such as medical fraud analysis, insurance claim analysis and money-laundering.

Making that data tangible, even visceral, is Buenen’s mission.

What a data network looks like.
What a network looks like: SynerScope makes it possible to pick trends out of huge data sets at a glance.

A trained architect who has spent a career massaging and managing corporate data, Buenen founded SynerScope to make the data coursing through governments and enterprises more visible. “The screen we look at this data through has always been one with 1,000 or so data points,” Buenen says. “If I can make a screen that holds 100 times, 1,000 times more data, without people getting lost, I can offer people a vision into big chunks of raw data.”

It’s a data visualization technique perfected by Danny Holten, SynerScope’s chief scientist, who Buenen met at the University of Eindhoven, a global hub for data visualization expertise. The 11 employee-startup turned Holten’s academic work into an enterprise-class product earlier this year. So far, SynerScope has three customers – the U.S. National Insurance Crime Bureau, Dutch insurer Delta Lloyd and a large American insurance company.

At the center of Buenen’s product is a seemingly simple visual metaphor — a series of concentric rings showing connections between one data point and another. These allow users to quickly sort through massive quantities of data. But that simplicity is deceptive.

To demonstrate, Buenen loads up a huge data file documenting the activities of a 20,000-node botnet, a network of infected computers used by hackers to steal credit card information and attack big web sites.

Within moments, Buenen’s tool can be used to identify who is talking to whom, where, and when. The information is all there – hidden among 173,000 rows of raw data. SynerScope just provides the visual metaphor that makes it all comprehensible, quickly enough to act on. “See that one, that’s where the police are going to want to go,” Buenen says.

The graphical-user interface, he asserts, does away with the need to master a complex query language. Buenen calls the result “data analysis for the rest of us,” echoing an early Apple slogan.

To make it work, however, Buenen’s product needs a dose of visual computing muscle. “Our product wouldn’t function without NVIDIA,” says Buenen, who has to pull out a laptop equipped with an NVIDIA graphics chip to perform his demo.

His next project: adding the ability to render data on servers equipped with NVIDIA’s Quadro GPUs so the product can be used from a tablet computer. Or even an iPhone.