Context Relevant, Ayasdi Score Funds to Automate Big Data Analysis
[Clarified 7/18/13, 4:27 pm. See below.] Context Relevant and Ayasdi, two competitors trying to automate big data analysis, separately announced venture capital investments totaling more than $37 million Tuesday showing continued investor interest in the business of turning data into insight.
In Seattle, Context Relevant raised $7 million in a Series A round from Madrona Venture Group, Vulcan Capital, Bloomberg Beta, and angel investors including Geoff Entress. Madrona and Entress were among the investors in the company’s $1.5 million seed round last year.
Meanwhile, Palo Alto, CA-based Ayasdi attracted a $30.6 million Series B investment from Institutional Venture Partners, Citi Ventures, GE Ventures, Khosla Ventures, and Floodgate. The latter two had invested in the company previously. The company raised $10.1 million last year, and $1.5 million in 2010.
Both companies aim to provide their customers with the sophisticated big data analytics capabilities that would otherwise require an in-house team of hard-to-find experts in fields such as machine learning and distributed computing.
“What we’re trying to do is make it much easier for corporations and government agencies to benefit from the power of predictive analytics and machine learning without having to go hire a huge team of data scientists to answer the questions that the business leaders are trying to have addressed, nor should they have to go build specialized IT infrastructure to make it happen,” says Craig Fiebig, Context Relevant’s head of marketing. “Much easier, much faster, and much lower infrastructure costs are at the root of what we’re trying to do.”
Context Relevant gathers data from multiple datasets and applies machine-learning algorithms with “smart automation,” says engineering director Dustin Hillard. The system builds and tunes a model, a step normally left to data scientists. That model is updated to respond to changes in the data being analyzed and in the kinds of questions executives ask of it.
Ayasdi‘s approach centers on Topological Data Analysis, which organizes data as a topological network, the company says, uniting machine learning algorithms to find answers to queries. The company, which cites customers including GE, Merck, and Harvard Medical School, says its platform can be used by domain experts and business people “without coding, scripting, or manually querying.”
Fiebig says one way Context Relevant—which plans to introduce a refined version of its current, commercially available product later this year—will differentiate from competitors such as Ayasdi is with speed. [An earlier version of this paragraph suggested Context Relevant had not yet begun shipping its product. In fact, it is shipping a version now.]
“A lot of these data science tools will produce similar results in the end,” he says. “The question, really, is how long did it take for you to get from the time you first asked the question until you see an answer show up on your screen? Starting from bare iron, how long does it take until you’ve actually got a running functioning model?”
Fiebig says the company is targeting three industries at present: financial services companies, using it to continuously update the value of bundles of financial products; high-traffic Web sites, which use it to select relevant content packages to display to individual users; and online travel agencies, which are using it to anticipate inventory needs—three-star hotel rooms in San Francisco’s SoMa neighborhood, say—and present relevant offerings to potential customers.
Context Relevant, which has about 18 employees and counting, will use the financing to hire more engineers to deepen its behavioral analytics libraries for the three target industries, and for new industry verticals; and to put the finishing touches on the core software.
Pricing has yet to be announced, and will scale depending on the intensity of the analysis performed with the product.
Ayasdi says its financing round will support continued work on machine-learning automation; provision of public datasets that companies can use in conjunction with proprietary data; and a doubling of the company in the next year.