Three Boston Startups Finalists in Amazon’s $100K Web Services Challenge

Conditions today in Boston: increasing clouds. Cloud-computing startups, that is.

Out of the seven companies named Friday as finalists in Amazon’s latest AWS Start-Up Challenge—designed to reward the most innovative companies using the cloud-based Amazon Web Services infrastructure to deliver their products or services—three are based right here in the Boston area. (Sorry, Seattle.)

Each of the companies—Watertown, MA-based MedCommons, Waltham, MA-based Pixily, and Dedham, MA-based Sonian—has turned to the Seattle-based e-retail giant for all or part of the computing power, storage, and other services they need to support their own customers. In that sense, they belong to a new breed of startups: nearly virtual ventures whose Web-based services free other companies from having to buy, install, and maintain their own software, but who pass along most of their own computing chores to Amazon.

Using Amazon’s cloud computing infrastructure is a great way for any startup to reduce costs. If you don’t have to spend hundreds of thousands of dollars of your investors’ money buying servers, after all, you can spend more on hiring smart people, or just keep a larger war chest on hand. But the executives I interviewed at the three local Amazon finalists say there are plenty of other reasons to go the cloud-computing route. In fact, it’s likely that these three companies—who will compete in Seattle on November 20 for a $100,000 grand prize—made it to the finals at least in part because they’re putting Amazon’s services to such unique uses, with a variety of benefits for themselves and their customers.

Sonian, for instance, provides business customers with a hosted e-mail archiving system, so that companies trying to comply with Sarbanes-Oxley requirements and other regulations don’t have to worry about manually backing up the data in their corporate e-mail systems from Microsoft, Novell, IBM, Google, or other vendors. The company has never stored any of its customers’ data on its own premises: it all gets shipped off to S3, Amazon’s storage service.

“Some of our investors, back in 2006, were skeptical of our ability to launch this service from the cloud, but we made the right bet,” says Greg Arnette, Sonian’s CTO. “Cloud was the way to go. It’s allowed our company to adopt the same philosophy that Amazon Web Services adopts around the pay-as-you-go model, so you’re not paying for hardware you’re not using. Also, even though we’re a new company in the space, our customers know their data lives on the infrastructure of a well-known, trusted brand.”

Adrian Gropper, chief scientific officer at MedCommons, says working with Amazon also eases fears about scalability. MedCommons allows healthcare consumers to create a central, online storehouse for their medical records; selected caregivers can download or upload records to this storehouse, including big image files like CT and MRI scans. “A lot of people can go into the health-record business and say they have a system that scales to 10,000 patients, but having it scale to a million patients is not quite as easy,” says Gropper. “Amazon provides us with a crucible. We haven’t proved that we can scale to a million patients, but we are using technologies that are compatible with that.”

There’s another big reason to use Amazon’s system as the infrastructure for a personal-medical-record system, says Gropper: it’s not Google or Microsoft. Both of those companies are developing online health record systems, under the Google Health and Microsoft HealthVault brands, respectively. But hospitals and physicians’ offices, which must verify the privacy of medical records under regulations such as HIPAA, may be unwilling to work with Google or Microsoft, who don’t let outside organizations review or modify the code underlying their systems. “Under these new ways of practicing, physicians and hospitals are going to need to be in control of their privacy policies, and the terms of use with either Google or Microsoft are not necessarily compatible with that,” says Gropper.

Pixily, which we profiled in October, calls on Amazon’s web services in yet another way, renting both processing and storage as it scans consumers’ paper documents and places them in a searchable online archive. Co-founder Anand Rajaram says that for Pixily, the great thing about Amazon’s computational utility, the Elastic Compute Cloud (EC2), is that more servers can be requisitioned as needed—say, around tax time, when Pixily is scanning and doing optical character recognition on peak volumes of documents—and just as quickly relinquished. “We don’t have to commit a lot of up-front capital for those five or six times a year when things spike,” Rajaram says.

Rajaram says that considering how small Pixily still is, Amazon has been remarkably responsive to its needs. “When they were here meeting with startups in October, one of the pieces of feedback we gave them was that they should introduce tiered pricing for S3,” he says. “About two weeks later they announced exactly that. It could have been happenstance, but we like the fact that they are actively listening to feedback.”

The AWS Start-Up Challenge itself, of course, is a big part of Amazon’s effort to reach out to the community of potential cloud computing users. Rajaram says it was attending the Amazon road show previewing the first AWS Start-Up Challenge, in 2007, that convinced Pixily to go with the company’s cloud services. And now, just one year later, Pixily is in the finals—competing with MedCommons, Sonian, and four other startups for $50,000 in cash and $50,000 in AWS credits.

Even if the company doesn’t win, Rajaram says, “it will be a platform for us to be endorsed by somebody like Amazon, given that they are the pioneers in cloud computing. More than anything else, it’s a big PR win.”

Wade Roush is the producer and host of the podcast Soonish and a contributing editor at Xconomy. Follow @soonishpodcast

Trending on Xconomy

By posting a comment, you agree to our terms and conditions.

  • NomaDesk’s features are based on a Virtual Fileserver Network (VFN) that stores the files you share on your local hard drive and our central server. This scalable, distributed file system is functionally transparent to you and your team, offering any “invited” user secure, seamless and effortless access to documents.