Mapping the Cloud With Ann Arbor’s DeepField Networks
When I was talking to Craig Labovitz last week about his new big data startup, DeepField Networks, there were moments when I wished the Internet really was just a series of tubes. That would probably be easier to understand than the massive, many-headed beast that it is today, which is why Labovitz launched his company in the first place.
The cloud has changed everything, he says, and he’s certainly in a position to make that call. He was the chief scientist at Arbor Networks, the Chelmsford, MA-based startup that grew from being a DARPA-funded research project at the University of Michigan to the traffic management and security tool used by 70 percent of the world’s Internet providers. Before that, Labovitz spent more than a decade working for companies like Microsoft Research and Merit Networks.
“The way that large Internet providers build infrastructure and monetize it is all changing,” Labovitz explains. “There’s a massive disruption in how money flows. It used to be simple, but now it’s complicated. Everyone is building a service, and there are lots of different components and players.”
What Labovitz and his team of engineers do is create “cloud genome” technology. DeepField—which boasts some staff members who worked on the original “pre-Internet” NSAnet— analyzes huge volumes of cloud and networking data, providing real-time cost analytics. Having this data, Labovitz says, enables providers to launch new services quicker, optimize cost, and improve the performance and availability of their cloud infrastructure. “I don’t think it’s ever been done before—the mapping of the public and private cloud,” he adds. “We’re not just looking at web pages, but how all the pieces fit together.”
Of course, to you and me, all this might seem abstract. After all, we just type in our URLs and wait for the magic to happen without much thought about what’s going on under the hood. But that’s precisely where the revolution is happening, as big content providers like Netflix and Amazon dominate the market and control an ever increasing amount of traffic. “If you’re a large enterprise with multiple data centers or millions of subscribers, you’re suddenly dealing with a very different world,” Labovitz says. “You’re probably working with 20 or 30 companies that control content distribution. You’re pulling together a best-fit toolbox.” Imagine the Internet industry supply chain as being similar to the auto industry or the railroads, Labovitz says. What DeepField wants to do is improve efficiency.
DeepField Networks, which is backed with capital from DFJ Mercury and RPM Ventures, has been around less than a year. Labovitz says it already has its first paying clients, though he declined to name them other than to say they are “large content players.” The company plans to scale rapidly and is hiring. “A lot of the story up to now has been about connectivity and access,” he notes. “What’s exciting to me is the massive amount of data and real-time access to that data in the cloud. The tools I have are just incredible compared to a few years ago. It’s such a fascinating evolution.”