Following your interests and making connections can launch a career.
Unlike most US students, Nathan Sanders declared his specialty as soon as he started undergraduate studies. He’d known for years that he wanted to study astronomy, but during his PhD at Harvard University he realized that the analysis itself enthralled him more than the applications for astronomy. He describes how he used his technical skills, and connections outside his academic program, to launch his career.
What caused you to move into data science?
When I was completing my PhD, data science was starting to be a big buzz word, and to be embraced in various industries. I could recognize that commercial groups were looking for people with the skillset I was developing. I decided to take on computation as a secondary field. I knew I could try new statistical tools in many fields and not just astronomy.
Did you seek experience outside your academic program, and did it influence your job search?
I hear again and again that the best place to start out with a job search is with the connections you already have. For scientists, my advice is just to be an active member of your community and to interact with people outside your direct research subject.
I wanted to do data analysis outside of academia, and I’m passionate about environmental issues. I took a summer fellowship through Harvard’s Kennedy School of Government in the Massachusetts statehouse, which was a phenomenal opportunity. As a scientist, I could do an analysis of sewage overflows in the Mystic River watershed and to use GIS tools that a legislative office wouldn’t normally be able to apply. I made these maps and walked around the statehouse arguing that municipalities should have to notify the public when certain discharges occur.
Also there are two projects that I got started as a graduate student that did science outreach and communication. One was ComSciCon, which empowers young scientists to be ambassadors for their fields. One random connection led me to Legendary Entertainment, where I work now. It’s a good example of how job -matching is generally done in the corporate world, which is through networking rather than the traditional method of trawling a job board online or doing the career expo circuit.
What is the biggest difference in workstyle between your academic and corporate positions?
The biggest shift by far was one of time scales. In academia, I was used to working on projects over a year. Projects that I work on today are likely to take weeks, possibly even days. Every once in a while there is a deadline of just a few hours. That is such a difficult transition to make. When there is a quick deadline that means we can’t explore all the aspects of data we’d like, we just do our best for that first time. We realize that this problem will come up again, and we’ll plan to do even better the next time.
You’re now in a position to hire data scientists yourself. What qualities do you look for?
There are three skill sets that we look for. One is the baseline technical competence to do the data science work required. It’s pretty easy for us to screen for the technical abilities from the resume.
The next two categories we can discriminate in the interview process. One is communication; it is so crucial. We need people comfortable interacting with the production side – filmmakers – and equally comfortable interacting with the business and strategic consulting side. And the third thing that we look for is a level of independence and free thinking. I think PhD students really should have the ability to demonstrate that in spades. There’s not a legacy of literature on how to do data analysis in the movie industry. Graduate students innovate and lead their own projects and learn how to come up with a new solution rather than rely on an existing method.
Interview by Monya Baker
For more, see Trade talk: Star selector