Dato, Formerly GraphLab, Raises $18.5M for Intelligent Apps

[Corrected 1/9/15, 2:43 pm. See below.] Seattle startup Dato, formerly known as GraphLab, has raised an $18.5 million Series B round to continue building tools that allow companies to make predictive applications that use machine learning to parse enormous databases.

The funding comes from new investors Vulcan Capital, which led the round, and Opus Capital Ventures, as well as previous investors New Enterprise Associates and Madrona Venture Group, which co-led the company’s $6.75 million Series A funding round in May 2013.

Co-founder and CEO Carlos Guestrin says technology giants have employed teams of data scientists and machine learning experts to build intelligent apps that make predictions and recommendations and segment customers by analyzing vast amounts of data. Dato wants to help other businesses that have invested in collecting data, but don’t yet have the capabilities to make useful predictions with their data, do things like identify customers who are likely to leave and take actions to keep them on board.

“It’s a time of great innovation in this space,” Guestrin says. “Companies like Google and Amazon have been able to do great things, but now everybody who’s put their data to one place and done some basic statistics wants to do more with their data.”

Steve Hall, managing director of Vulcan Capital, says the company’s technology is effectively a force-multiplier for sought-after data science experts.

Dato is filling a huge market gap by bringing together the ease of use and computing scale that make it possible for one data scientist or developer to do the job of many,” Hall says in a company news release.



Guestrin says Dato’s customers include Zillow, Adobe, PayPal, and Cisco. The company has users numbering in the thousands, with its active users growing about 50 percent month over month, he says.

The new funding will support continued new business development, engineering, and customer support. Dato has 27 employees working from its headquarters in the Fremont neighborhood and plans to add employees on pace with the growth of its users.

The name change reflects Dato’s focus on data of all types. It’s one of several Seattle-area technology companies that make up part of an emerging “dataware” technology stack.

GraphLab was a holdover name from the open source, academic research project begun about seven years ago at Carnegie Mellon University that gave rise to the company. That research focused on machine learning applied to large-scale, multi-dimension graph datasets, in which nodes represent objects—such as people, pictures, and other entities in a social network—and edges represent the relationships among nodes, such as likes and tags.

“As we brought GraphLab Create to the market and we looked back at the name that was originally created as an academic project for graphs, we realized that it didn’t capture where the company is, and where the company is going in the future,” says Guestrin, who spoke at Xconomy’s November 2014 forum on innovation in Seattle.

The software today handles data in an array of formats from tables to text to images. “Being able to build a product that combines all this kind of data was something really exciting to me,” he says.

Guestrin and colleagues settled on Dato, which is Spanish for datum—a piece of information. “It’s really simple and has lots of potential,” he says, noting that the company considered a lengthy list of nonsense neologisms that tech startups often hang on themselves. The name also nods to Guestrin’s own heritage: He is from Brazil by way of an Argentinean family and speaks both Portuguese and Spanish. [An earlier version of this paragraph mistakenly stated that dato is Portuguese and Spanish for datum. The word in Portuguese is actually dado.]

Trending on Xconomy

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

2 responses to “Dato, Formerly GraphLab, Raises $18.5M for Intelligent Apps”

  1. really bilingual says:

    “Dato” means datum only in Spanish. In Portuguese it’s called “dado”.

  2. Indeed, a quick check of Google Translate, confirms your correction. Thanks, really bilingual.