"It is semi-structured data, in that you might have information on who posted it and the date, but the bulk of it is a load of text, so you have to teach the computer to spot the risk in all of that unstructured data, and that is where the natural language processing comes in."
He added: "We have spent our time building something that can very quickly cluster large amounts of information into 'interesting', 'not interesting' and 'very interesting' buckets, so that we can cut down the amount of time that is needed to spot the risk in all of that information."
While there are already social media monitoring platforms used by banks, performing these tasks individually is a human-intensive process, Paterson said, meaning that the ability to provide an overview of a wide variety of threats has caught the attention of major financial organisations.
"These banks are being attacked every day, so they are interested in any information and tools that can help them," he said.
Digital Shadows was one of seven companies involved in the FinTech Innovation Lab start-up accelerator, and presented in front of major banks as part of an investor day yesterday, having already picked up innovation awards from Innotribe and Cisco for its software service.
Speaking at the investor day, Paterson told Computerworld UK that changes in the financial sector have meant that way banks are beginning to reconsider how they develop systems.
"The interesting thing with the financial crash is the banks can no longer sustain huge internal development teams trying to build it all themselves. They have got to look more to smaller companies on the outside which can give them specialist services," he said.
"So, in some respects, it is easier now than ever to get through to the banks, as they are having to change and innovate with the environment."
Sign up for MIS Asia eNewsletters.