Trifacta raises $12M to help make dirty data clean

Trifacta, a San Francisco-based startup focused on turning raw data into usable data, has raised a $12 million series B funding round from Greylock Partners and Accel Partners. The company uses techniques from fields including machine learning and human-computer interaction to save users the time and effort often required to make their data usable by most business intelligence software or other analytic programs.

Traditionally, transforming data, even at a small scale, is a cumbersome process that often takes much longer than the actual work of analyzing it. This problem is exacerbated in a big data environment, where there’s more data to work with and where new formats can make it harder to clean up. Trifacta’s software is currently in use by beta customers, Co-founder and CEO Joe Hellerstein said, and he’s seeing three big use cases emerge: business analysts who want to work with data in BI tools; IT personnel whose job it is to transform data; and data scientists who want to work with data in programs like R, or marry new types of data (like streaming data) with existing data.

Trifacta’s approach — like that of fellow Accel investment Paxata, which recently launched — is to let users work together with the software to teach it what they want done with specific data. For example, data from certain sources might always be transformed one way or ported somewhere else, certain rows of data might always represent specific values, or certain datasets might often be joined together. Trifacta has strong relationships with companies such as Cloudera and Tableau, which represent popular beginning and endpoints for the data Trifacta is used to format, Hellerstein said.

Trifacta’s founders include Hellerstein, a renowned computer scientist and University of California, Berkeley professor; Jeffrey Heer, a University of Washington professor and former head of Stanford’s visualization group who helped create popular tools including D3.js; and Sean Kandel, a former financial analyst and Stanford Ph.D. candidate studying data manipulation. The company launched in October 2012 along with $4.3 million in funding from Accel Partners and a handful of seed investors, and a who’s who list of advisers.

Feature image courtesy of Shutterstock user Alexander Fediachov.