Gregory T. Huang
Big science and big technology work hand in hand. Without the World Wide Web, an idea conceived by scientists at CERN (the European Center for Nuclear Research), we wouldn’t have technologies people use every day, like social media, e-commerce, or Internet cat videos. And, as my colleague Wade argued, without the Web, scientists probably wouldn’t have garnered the necessary support for CERN’s Large Hadron Collider, the gigantic facility at which the Higgs particle’s existence was more or less confirmed this month (no, I’m not going to explain what it is, but you can see more about it here).
Now a Boston-based startup could represent the next phase of technology advancement. Cloudant, which also has significant operations in Seattle, provides what it calls “database as a service.” The company talks about building a distributed “data layer” around the globe. None of that quite does justice to what it is doing, however.
The connection to big science here is that Cloudant’s three founders are particle physicists from MIT. Alan Hoffman and Adam Kocoloski did research at Brookhaven National Laboratory, and Mike Miller worked there and at the Large Hadron Collider. These guys were cranking on “big data” before that was even a term. In 2008, they started Cloudant to commercialize database technologies inspired by their experience wrangling some of the biggest datasets in the world.
“We spent too much of our time on managing analysis jobs, finding the data, data provenance, bookkeeping, and overhead,” says Hoffman, who served as Cloudant’s CEO until late last year. “We were experts, but still having trouble keeping track of information. The tools were not very good.”
Along their path to starting the company (and finding its focus), the Cloudant team gained admission to the Y Combinator accelerator program in the summer of 2008, in Boston. (There is some confusion on the Web with another YC company, SlapVid, which changed its name to Cloudant in 2007 but is unrelated.) Hoffman says he took away one piece of advice in particular from Paul Graham, the program’s head: Be unkillable, like a corporate cockroach.
Fast forward to today, and the company has 22 employees—10 in Boston, five in Seattle (including co-founder and chief scientist Miller, who was also a research professor at the University of Washington), and the rest distributed. Cloudant has raised a total of $4 million from Avalon Ventures and Y Combinator. Last November, it brought in Derek Schoettle, formerly an executive at Boston-area data analytics firm Vertica Systems (now part of HP), as CEO.
Database companies, by their nature, are pretty technical beasts. Without going into mind-numbing detail, what Cloudant does is provide a data management network for developers of Web and mobile software. Cloudant’s product is based on an open-source project called Apache CouchDB, and it’s “built specifically for the distributed nature of the cloud,” says Hoffman. That means it has to be fast, reliable, and able to scale up from handling data from one server to hundreds of servers. To that end, Cloudant runs a network of sites from the U.S. to Europe to Asia. “We shrink-wrap a data layer around the globe and you tap into it,” he says.
Cloudant competes with Amazon’s DynamoDB product (part of Amazon Web Services), as well as any number of smaller companies offering scalable, cloud-based database services. But the startup is really part of a larger movement of companies looking to disrupt entrenched database giants like Oracle and Microsoft. Another way to slice it: what Akamai did for Web pages and static content, Cloudant wants to do for “dynamic data,” Hoffman says.
Hoffman wouldn’t divulge any company revenue stats, except to say the past three quarters have been “really exciting,” and that the market has come around to Cloudant’s point of view. He adds that the company has more than 8,000 small to mid-sized customers (who pay $500 or less a month) and about 50 large customers who pay much more than that.
Cloudant seems to have found a niche across developers of Web, mobile, and tablet apps, and across analytics and gaming software from small startups to big enterprises. But the team is wary of being stretched too far, too soon.
“We’re trying not to be all things for all people,” Hoffman says. “We’re very focused on that thin but vital layer that is the data.”