By Allan Maurer
CHARLESTON, SC—Electronically scouring unstructured data to discover anomalies or connections has quite a lot of potential for various purposes, but BeliefNetworks plans to focus on two primary but unconnected areas, disaster and emergency support and ad networks, says CTO Ted Tanner.
Both Tanner and his co-founder, Lisa Maki, president, have worked with major players in the IT industry and have specific experience in data mining. At Microsoft, Maki helped define a new product in the area of International Disaster Response and Recovery for the company’s Research & Incubation team, with an emphasis on knowledge discovery and data fusion.
Tanner was a systems architect for Apple and an architectural strategist for Microsoft.
He also played instrumental roles in four start-ups, including Avid/digidesign, Crystal River Engineering (bought by Creative Labs) and Mongo Music, which sold to Microsoft for $75 million.
Harvests and analyzes data
BeliefNetworks has built a solution for harvesting and analyzing all relevant data, structured and unstructured, in a given situation. The software fuses numeric data analysis such as machine learning and data mining, with symbolic methods such as semantic intelligence.
The software continuously learns and adapts, optimizing for events and actions most likely to occur in the future and the information exchange is authenticated to preserve data integrity, security and business rules.
The nine-employee company has raised about $1 million in angel funding and recently reopened its seed round. It would like to raise from $3 million to $6 million, Tanner says.
Company is hiring
The company is hiring people in distributed computation, machine learning, and a senior software engineer.
“We built all of our tools without external code,” says Tanner. “It’s production level software development.”
In the public safety arena, the software takes highly unstructured information and applies predictive analytics. It will map data to concepts to find anomalies.
Predicts which ad will do best
“Our predictive analytics platform identifies anomaly to expose known and previously unknown or unexpected areas of risk,” says the company’s executive summary.
For ad networks, the software can analyze a publisher’s page and tell the network which ads will perform best, says Tanner.
The system crawls the publisher’s web sites, performs semantic intelligence on the content, and matches concepts in content to content in ads. “We’re able to predict the best ad for the publisher in real-time,” Tanner says.
“We’ve seen significant success in increasing click-through rates,” he adds.
Four products developed
The company has developed four products. DataChasm is a knowledge fusion engine that collects, cleans and analyzes data to identify previously unknown relationships.
AffinityAgent makes relevant recommendations for the distribution and placement of content and ads.
PreparednessAgent is an automated emergency alerting and contact management system. DynamicDataAgent offers a negotiated exchange of structured information at the semantic level.
The company has implemented its software in the Amazon cloud computing environment.
“It’s beautiful,” says Tanner, “and reduces costs. So we have an extra-low burn rate. We do have an internal computer cluster where we model things before putting it up on the ether.”
East Coast, West Coast differences
He says that while the company has been approached by healthcare providers, it is not ready to engage in that sector yet.
“Right now we have to pick our battles and we’re focusing on public safety and ad networks,” he notes.
Tanner says starting the company in Charleston rather than Silicon Valley has advantages. “People said we wouldn’t be able to find the right people. But I have stacks of resumes to choose from.”
Not only that, he says, the company is also working on developing a perceptual model of information and mapping that to parameters on which to compute.”
That could help create a dynamic personal portal.
“That idea appeals to people. People are tired of leaning forward to use the Internet. They want to lean back and be entertained, like watching TV. Several heavily funded startups on the West Coast are gunning for this. But our entire burn rate is probably equivalent to what their office space costs in a month.”
Tanner, who spent a considerable amount of time in Silicon Valley says they do look at the startup investment scene somewhat differently than on the East Coast. “On the West Coast they’re more concerned with the team and not necessarily the idea,” he says.
“They know a great team can turn a B idea into an A idea. It’s all about execution. In the Southeast, there is more emphasis on customers and revenue, and there, an A-round is akin to an angel round here.”
On the Web: www.beliefnetworks.net
© 2008, TechJournal South. All rights reserved.
