Between allegations that they manipulated benchmark interest rates or wrote faulty mortgages, banks have racked up more than $170 billion in litigation costs since 2008, according to an estimate by the Macquarie Group (MQG:AU). So some financial institutions are investing in technology that can sift through millions of e-mails, instant messages, and transcripts of telephone calls to spot suspicious behavior before it explodes into a mess of lawsuits and fines.
“What the banks want is a way to surveil and read all these things as they happen,” says Tim Estes, chief executive officer of Digital Reasoning, a data analytics firm that on Oct. 8 announced a $24 million investment from a group of banks led by Goldman Sachs (GS) and Credit Suisse (CS). “If you’re working on a mergers-and-acquisitions team at a bank and giving information to a trader, that ought to be flagged when it happens, not in an investigation three years later.” (Investment banks are supposed to have so-called Chinese walls between staff making investment decisions and those working on deals.)
Digital Reasoning, founded in 2000 in Franklin, Tenn., specializes in software that takes into account the context of written communications—for example, knowing when a person is referring to apple the fruit vs. Apple the stock. As it works its way through the material, the program builds a store of knowledge that it can delve into to map relationships between individuals or events—a process called machine learning. The U.S. government has been a Digital Reasoning customer since 2000, says Estes, and In-Q-Tel, a venture capital firm affiliated with the CIA, is an investor.
Read the full article here.