The idea of recording the future sounds like something from a science fiction flick, but that’s, in a way, what one Somerville-based company is working toward.
The startup’s moniker says it all: Recorded Future.
“These days, we do live up to our name quite nicely,” said Staffan Truvé, the chief technology officer and co-founder of the cyber-threat intelligence company.
The late-stage startup uses specialized software to sift through tens of millions of documents every day in 10 languages including Russian and Chinese dialects, scouring them for clues of future threats to companies and other organizations.
Founded around nine years ago, Recorded Future has so far raised $57 million from investors such as a Google and In-Q-Tel, and in October reeled in a fifth round of funding – $25 million – led by Insight Venture Partners. The company has also grown to more than 200 employees, and has expanded to locations in Sweden, Washington, London and most recently Singapore. Its customers include big names such as Fujitsu, T-Mobile, St. Jude Medical and Japan’s NTT.
As the company has grown, their strategies and products have also evolved, said Truvé during a phone interview last week from his native Sweden.
“In the early years, the way we predicted the future was purely based on the way that people talked about the future. It was based on people talking about planning future activities: You had a ... group planning to hack a web service. … They used the Web to coordinate,” he said. “But these days we also have some truly predictive stuff.”
Algorithms, analytics software and other systems help them figure out trends and potential threats, he said, as they penetrate the so-called dark web and elsewhere. Much of the work is automated – the data-gathering and analysis, for example – while some of it requires the eyes and minds of staff who weed through anomalies, connect dots and toss flawed data.
“We’re constantly broadening and deepening the coverage of data that we get – everything from big media to blogs to essentially the very dark web – hacker forums and that kind of thing,” he said. “If you collect enough information about what’s being written about the world, you should be able to model it and use that data to give interesting predictions.”
BostInno recently spoke with Truvé and the company’s chief executive, Christopher Ahlberg, on two different interviews about the company, its funding and approach to intelligence.
BostInno: What would be considered a threat?
Ahlberg: I’ll give you three kinds of threats, depending on who you are. If you are a journalist, for example, you could be worried about intelligence agencies... criminals or hacktivists who don't like what you're writing about... If you are writing pointy stuff, they might try to steal your what you’re writing about, they might try to scare you, they might try to influence you. If you’re a company, you might be worried about intelligence agencies who are trying to steal your intellectual property. If you’re a bank, it might be criminals who are trying to steal your money…. If you’re (a government), you might be worried about other intelligence agencies who are trying to steal your information.
BostInno: Could you give an example?
Ahlberg: ...It might be in a criminal forum, very deep in the underground where the bad guys sell and share and collaborate. There might be a guy who says, “Look, I have logins for Bank XYZ, does anyone know how we can take advantage of that?” Another guy might say, “That’s interesting, because I have a trojan that works exactly for that bank. I’ve been looking for their logins.”
BostInno: How do you define the dark web?
Truvé: My definition is the part of the web that is not indexed… by search engines. That’s partly because it’s hard to find and that’s partly because it’s protected by passwords and things like that. Say you set up shop in the dark web, you need customers to find you, you need to advertise. Essentially what they do is, they advertise – they might be in “less dark” forums, and what not, advertising their services – and then we can follow them and that’s the way we can find them. If you do business, you need to be found.
BostInno: So is this different from what a government might do, where they might put out a wide net, looking to, say, prevent a terror attack?
Truvé: No, we do cast a very wide net. … Any company could walk in the door, and we would have interesting information for them, because we do such a broad collection. So the comparison of how a government runs its intelligence agency is a very good one. I sometimes say that we are the intelligence agency for people who can’t afford to buy their own or have the technical skills to do their own.… My personal ambition is to build a model of what’s happening in the world… and then organize everything for analysis.
BostInno: Is there a place where you feel your client-base is the strongest, or is it more or less the so-called “Western World”?
Truvé: That’s a good description. Some people think we have a lot of government customers. We have some key government customers, but the bulk of our customers are large corporations, world-wide. It’s spread across many different sectors: finance, manufacturing, food production, infrastructure, all over.
BostInno: Is there concern someone could use this for bad purposes? What if, say, the Chinese government wanted to use your product to protect from people organizing against a political figure?
Truvé: We have very high moral standards. What you said earlier about selling to the Western democracies is a strong principal to us. The Chinese government has never contacted us and if they did we wouldn’t sell to them. With all the analysis (of) problems about China, I’m sure they would not contact us.
BostInno: You just raised $25 million in October. Do you expect to raise more at some point?
Ahlberg: I hope not. It’s not in our plans to raise more money.
BostInno: Do you have any plans for an exit?
Truvé: We want to build a great company. If we focus on building a good company, everything else will work itself out one way or another.
These interviews were edited for brevity and clarity.
