Future Point Tackles Data Issues with New Starlight Release

Aug 25, 2009


      Bookmark and Share

Some of the biggest and most twisted information knots lie in the data banks of the world's largest organizations: Fortune 500 companies and the American armed forces. Future Point Systems, a small, privately-owned company from Mountain View, Calif., has set out to untangle those knots with its flagship program, Starlight Visual Information System.

Starlight is already in use, boasting between 50 and 100 clients, FPS CEO Ben Sommers says. But a new version, released Wednesday, August 19, could change the way users aggregate and absorb data. "It's really a major release for us," Sommers says, though he was reluctant to speak about how the tool actually works.

FPS began developing Starlight in the mid-1990s, when it was still part of Pacific Northwest National Laboratory (PNNL), one of the U.S. Department of Energy's research facilities. Since striking out on its own in 2006, FPS has maintained close government ties. Sommers estimates that roughly three quarters of Starlight users are government entities, mostly intelligence, military, and law enforcement organizations. These organizations have turned to Starlight as a way to aggregate large amounts of data from many different sources, a concern of paramount importance when the volume of data that needs to be sorted is high and time is short, as is often the case in military or law enforcement situations.

To Sommers, increasing efficiency and speed are two of the biggest selling points of Starlight. "This is not a trivial part of the process for people using vast disparate sources of data," Sommers says. And the increase in efficiency can be striking; according to Sommers one of FPS' military clients spent 3 months analyzing one set of data, then, after switching to Starlight, sorted the same data and got similar results within a matter of hours.

Starlight is aimed at organizations that need to digest large amounts of data coming in from multiple sources quickly. Where the data originate depends on the user-it can come from internal records or external sources, and depending on time and space constraints, can be plugged into the system either manually or automatically. These data can be anything from text, to numbers, to geographical information, as FPS customizes Starlight for each client.

Starlight works by scanning this information and grouping certain points with similar ones, mapping them on a visual interface that looks, for better or for worse, like a chart from an astronomy textbook. Nonetheless, according to Sommers, the visual interface is one of the most appealing aspects of Starlight, allowing users to see data mapped out instead of having to sift through mountains of data and finding correlations themselves. Instead, the visual interface offers users a quick view of the relationships between data points and can show different levels of detail, from a big-picture overview to a more detailed perspective.

"Even though the visuals might look a little intimidating, it's very intuitive," Sommers says, adding that some users have been able to jump onto the system without any training, while most others don't need more than 2-5 days.

The new version of Starlight will retain the key features of the original, but will be geared more toward making it easier for clients not only to collect and analyze data, but to disseminate it. While older versions of Starlight did not include tools to view data outside the system, the newest release will offer one-click reporting, speeding up the flow of information from analysts to higher-ups. The one-step reporting, which Sommers characterizes as "not quite web-based," means that users no longer need to use Starlight to view results generated by the program.

For data that needs to be sorted by geographic location, the new Starlight will now support Google Maps, though Starlight will continue to run FPS' proprietary mapping system.

Sommers thinks the new features will take some of the burden off of analysts by making the functionary parts of their jobs easier. "It allows them to import data more quickly and do what they're really good at, which is the analysis," he says. Once the data are aggregated and sorted, analysts can figure out which points are important and how they are related to others.

Sommers hopes that the release will lead to a new crop of users, both inside the "core market" of military and intelligence organizations and in commercial circles. "I think it will absolutely allow us to expand with our core market," Sommers says, "but then on the commercial side, the answer is even more emphatic, that consumers need to access and aggregate data quickly."

(www.futurepointsystems.com)