����ֱ��

In the wake of the 2008 crisis, “the world of finance was broken,” states Stephen Piron, a computer science grad and entrepreneur. Many governments tightened regulations, but the problems persist – albeit on a smaller scale, he says.

Piron, who graduated with a BSc in 2003 from Victoria College at the University of Toronto, takes a keen interest in the state of the world’s financial system. He’s working upwards of 100 hours a week running Bright Sun, a data analysis company he founded in London two years ago with the notion of helping financial companies stay on the right side of regulations. (.)

“The idea of using data analysis for fraud detection isn’t new, but a series of bank scandals created a bigger appetite for cutting-edge surveillance tools,” says Piron.

Both the banks and the regulators grew fearful of the reputational damage that additional scandals could cause, and Piron recognized a market opportunity.

“There is now a huge appetite for tools to pick out fraudsters.”

Banks contend with vast amounts of data, related to thousands of transactions their traders carry out every day. The challenge they face, along with regulators, is finding instances of abuse.

“Banks have a duty to make sure their traders play by the rules,” says Piron. “We provide the tools for policing.”

He says the organizations using Bright Sun’s software, which also include hedge funds, are alerted to potential abuse as it happens, allowing senior managers to investigate quickly to determine if abuse actually occurred. Managers can also look back over historical data for incidents with similar traders in the past. In other words: real-time and historical fraud analysis.

The kind of analysis that Bright Sun offers – poring over buying and selling transaction patterns – has an additional use: it helps the bank determine if it has the right “inventory” of financial instruments at any given time. This sort of intelligence – the kind that enables a bank to know, for example, that its German clients typically buy Japanese government bonds at the beginning of the month – traditionally has been gathered by human traders through experience.

“Sometimes this human intelligence is excellent, but it is also subject to all of the biases humans are prone to,” says Piron. “We’re using a computer to gather the intelligence systematically and comprehensively.”

Bright Sun is not Piron’s first tech startup. In 2004, he launched SimonSays Voice Technologies, which he describes as one of the world’s first speech-recognition technologies for search engines. The software converted audio and video into text transcripts, allowing search engines to search multimedia files. He sold the company in 2007 and worked for six months on Bay Street in quantitative finance before shipping over to London, where he landed a position with Man Investments.

Then along came the financial crisis – and a market opportunity. Bright Sun launched as a solo venture, but now employs seven people full time; software engineers craft the data analysis tools while Piron handles most of the sales “through old-fashioned knocking on doors.”

Like his first venture, Bright Sun requires Piron to be on duty, or on call, 24 hours a day, seven days a week. It’s a position he’s comfortable with, he says, compared to working for someone else.

“In the corporate world, they put you in a box, and I don’t really fit in a box.”

Zoe Cormier is a writer with ����ֱ�� magazine. (.)