J Hellerstein

Joe Hellerstein joined the Berkeley faculty in 1995, after a PhD at Wisconsin and an MS at Berkeley. He is the Jim Gray Professor of Computer Science in the Department of Electrical Engineering and Computer Sciences.

"Historically, computer science treated data as the uninteresting input and by-product of computation. I have always seen that world differently: if you view data—your inputs and outputs—as the primary objective, this leads you to elegant, useful computer science." 

Research Focus: Data-centric computing, broadly cast: systems, theory, languages, interfaces, and applications.

How is this applied in the world? 

Work from my group catalyzed a new industry of "Data Wrangling." A widely quoted statistic states that 80% of the time people spend in data analysis and AI projects involves wrangling data into shape—acquiring, transforming, blending and cleaning data from multiple sources to get it ready for analysis. My group pioneered the design of intelligent visual tools for data wrangling that make these tasks accessible to non-programmers. Based on our research, in 2012 we started a company called Trifacta(link is external), which is now the leader in the global data wrangling market. Trifacta's customers include major organizations across many sectors, ranging from global banks and healthcare companies, to NASA and the Centers for Disease Control and Prevention. 

Why is this area important to you?

Data was the underdog in computing until the last decade. Historically, computer science treated data as the uninteresting input and by-product of computation. I have always seen that world differently: if you view data—your inputs and outputs—as the primary objective, this leads you to elegant, useful computer science. This "data-centric lens" on computation gave me the license to intellectually meddle in all aspects of computing from a unique perspective. As a result I have spent my career collaborating happily with computer scientists of many stripes across a broad range of ideas and challenges. In recent years I've turned outward, exploring the ways that the data-centric lens can help end-users of data and computation. This includes data "enthusiasts" in non-profits and business who want to use data to solve the problems in their domain, but do not care to become experts in computer science. 

Anything else you'd like to share?

I am an amateur jazz trumpet player. Over the years I have had the opportunity to perform with far more serious musicians who have gone on to great things. This includes current jazz heroes such as saxophonist Joshua Redman (an alumnus of Berkeley High who I played with in college), pianist Vijay Iyer (a former UC Berkeley PhD student who I played with in graduate school), and many others. During a recent data management conference in Los Angeles, I snuck off to record backing tracks for an LA-based Americana band called Great Willow that I hope will appear on their next album.