Follow Datanami:
March 19, 2021

Meet 2021 Datanami Person to Watch Joe Hellerstein

Alex Woodie

Joe Hellerstein is a busy guy. When he’s not working as the chief strategy officer at Trifacta, he’s teaching computer science courses at U.C. Berkeley or advising grad students at the prestigious RISELab. But we managed to catch up with Hellerstein, who is also one of our 2021 Datanami People to Watch.

Datanami: We’ve seen the emergence of interaction data completely swamping the scale of transactional data of the previous generation. What are you looking for in the next wave of data?

Joe Hellerstein: Interaction data from software has been a huge wave and it’s still washing over us. But right behind it is the wave of sensor data from the physical world. Think image and video, mobility and of course IoT. Until recently we mostly treated this as media, for archival, replay and possibly search. AI is now helping us extract structured “features” from media, translating physical signals into discrete data records that can be analyzed and tabulated. For better and for worse, we’re on track to literally record the physical world as bits and convert it to records—translating the raw bitstreams of sensing into higher-level, meaningful data for analysis. This is going to stretch our existing technologies, policies, and ethics.

Datanami: You’ve maintained your career in academia while simultaneously pursuing interest in private industry. Lots of other folks have had less-than-stellar success at that. What’s your secret?

Hellerstein: Success comes down to the people you spend time with. Primarily I’m lucky to have been surrounded by amazing mentors and colleagues, and it’s no surprise that many of them have also succeeded on both these fronts. The good fortune of being at Berkeley has been a big part of that.

In terms of suggestions, I’d lead with one of my favorite quotes: “Laziness in doing stupid things can be a great virtue.”

And anybody who does creative work needs to read Herman Melville’s short poem called “Art”. It has the prescription.

Datanami: How would you characterize the state of data science education at the university level? Are we bringing enough young adults into the programs, and are we teaching them the right things and exposing them to the right ideas to be successful upon graduation?

Hellerstein: It’s getting real, and we’ve been aggressive on this front at Berkeley so we’re navigating the edge of the wave. Our Data Science courses are now taken by thousands of students every single semester, and we’re starting to produce thousands of data science majors at Berkeley as well. Of course, it’s early and we’re still learning. One shift that I’m pushing for is more focus on Data Engineering. It’s not enough to teach statistics and machine learning algorithms on toy data in notebooks. We need to expose students to issues of wrangling messy real-world data and orchestrating complex scalable systems. I’m happy to say we’re offering our first undergraduate Data Engineering course at Berkeley this semester, and lessons from my time in industry have been very helpful.

Datanami: You’re from Berkeley, but your Trifacta co-founder Sean Kandel is a Stanford guy. Do you have an axe at Trifacta, and if so, who is holding it now?

Hellerstein: Yeah, well, we don’t need to go there. I mean, Sean definitely enjoys a little smack talk now and then. But I’m a lover, not a fighter. And as the grown-up, it’s my job to set the tone. Which is easy to do when the superiority of my institution is so obvious.

Datanami: Outside of the professional sphere, what can you share about yourself that your colleagues might be surprised to learn – any unique hobbies or stories?

Hellerstein: I play a lot of trumpet. One of my quarantine activities was to play the horn tracks for an upcoming album from James Combs called “Impolitic.” Check it out. On a somewhat less hummable front, I’ve also been participating in a weekly online free improvisation group called the duo B. Experimental Band (dBxB), led by Lisa Mezzacappa and Jason Levis. We’re simultaneously testing the limits of musical expression and the impacts of Internet latency on audio perception.

Hellerstein is one of 12 Datanami People to Watch for 2021. You can read the interviews with the other honorees at this link.

Datanami