From wine picks to stocks: Could the ‘big data’ geeks at newly-formed Context Relevant predict the future?

Stephen Purpura

Big data might very well be the tech buzz word of the year. And just how hot is it?

Consider this:  Data scientist Stephen Purpura — an expert in artificial intelligence, machine learning and predictive analytics who is studying for his PhD in information sciences at Cornell University — has received no fewer than 45 job offers in recent months. And they just aren’t any fly-by-night offers, with some rolling in from big-name companies touting salaries of $300,000 or more.

But Purpura, a former manager at Asta Networks and Microsoft who lives in Bellevue, has rebuffed all of the inquiries because he wanted to build his own startup. Today, the three-month-old Seattle startup, dubbed Context Relevant, is coming on the scene with a $1.3 million seed round of funding from Madrona Venture Group and angel investor Geoff Entress. Swype/Nuance executives Mike McSherry, Cliff Kushler and Mark Illing — who got to know Purpura while interviewing him for a job last year — also participated.

In addition to Purpura, Context Relevant was co-founded by Chris Metcalfe, a former marketing veteran at Isilon Systems who previously worked as a venture capitalist at Atlas Venture. Also on the team is Vice President of Engineering Jim Walsh, a 23-year veteran of Microsoft who previously led the company’s Cosmos data storage and analysis team. Darren Vengroff, chief science officer for RichRelevance; Mike Kail, vice president of IT operations at Netflix; and David Farber, a computer science professor at Carnegie Mellon University, are serving as a members of the company’s customer advisory board

Together, they are hoping to transform the way companies interact with the mountains of data they are collecting, making better sense of the data and drawing out new insights that previously would be difficult to discern. Purpura says they take the heavy-lifting associated with data analysis, and make it far easier, and quicker, to sift through the information.

“We employ a lot of very well respected data scientists, and the reason we do that is so our customers don’t have to,” said Purpura, adding that the company is currently working with a handful of undisclosed test customers.

Context Relevant’s cloud-based application essentially allows enterprise analysts and statisticians to function as big data scientists, said Purpura. In other cases, he said that it can help eliminate “data analysis paralysis” where companies “get stuck” capturing massive amounts of data, but aren’t quite sure what to do with it.

A bank, for example, could gain more insights from conversations on Twitter or a coffee company could use the tools to better understand how users interact with mobile coupons.

In one example from several years back, Purpura said he was able to use some of his sophisticated analysis to identify up-and-coming wine makers from northern California, buying up some of their wine before the prices skyrocketed in value. He’s since gotten to know the winemakers better, and they’ve helped introduce him to others in the industry.

“The software really made that happen for me. It helped me identify these key people, but I established the relationship. It is not like the software does magic,” he said.

If the software can help pick great winemakers before they are discovered, what about predicting something else of value, like stocks?

Purpura didn’t want to go into much detail on that, but he did say that “people on Wall Street are interested in the technology.”

It is extremely tough for companies to attract smart data scientists — since the best and brightest go to work with big pay packages on Wall Street or get quickly gobbled up at companies such as Google and Microsoft. Given that, Context Relevant wants to make it easier for those who might not be able to afford those hires to crunch data for their businesses.

Chris Metcalfe

“Our belief is that those (data scientists) are not going to be more plentiful as time goes on,” said Metcalfe. “So we are building those smarts into our templates, so that analysts can basically get up-and-running with that expertise, without having to find someone like Stephen or a team of Stephens to join them.”

Of course, Purpura could take the easy path and take one of those 45 job offers, and many in his position probably would. But he said there’s much more opportunity doing a startup.

“This is a fantastic opportunity for both me, and all of the members of our team,” he said. “The reality is that our skills can be applied across many companies, and it is likely that the company we build can generate substantial returns for our employees and our investors.”

As a result of the new funding, Madrona’s Tim Porter has joined the board. The company —  located in the “Darth Vader” building at 4th and Blanchard — employs 11 people.

  • http://twitter.com/Argentineanwine Argentinean Wine

    I am curious — which winemakers did he identify back them?

    • http://twitter.com/spurpura Stephen Purpura

      The modeling process identified two different winemakers operating in Napa Valley that people discussed online in a way that was indicative of strong performance that wasn’t linked to a extremely high Wine Advocate scores, each in a different category. One category turned out to be “balanced or feminine reds”, and the up & coming winemaker identified was Aaron Pott. Aaron is now the winemaker for Greer Wine, Pott Wine, and Seven Stones. The other category identified was “big reds” and the up & coming winemaker was Mike Smith. Mike had been an apprentice to Thomas Rivers Brown before he started out on his own. He now makes wine for Carter Cellars, Myriad, and Quivet.

      I stay in touch with both Aaron and Mike and it’s been fun to follow their careers. While I still don’t consider myself a wine expert, I enjoy using Context Relevant’s product to analyze data for my life. While it’s not the primary business purpose of the product, I will always be one of our product testers. And I enjoying sharing the fun insights that I can find in the data with my friends.