Dr. Chinmaya Gupta (CG) is a data scientist at Allstate Insurance Company. Prior to joining Allstate, he collaborated with molecular biologists on gene network modeling at University of Houston and worked on chaotic dynamical systems at University of Southern California. He obtained his bachelors in mathematics from St. Stephen’s College, Delhi and his masters and Ph.D. in mathematics and statistics from the University of Houston. Dr. Gupta blogs at https://chinmayagupta.wordpress.com/.
Mario Pinto (MP) had a face to face session with CG on his journey and his message for aspiring data scientists.
MP: What attracted you to a career in data science? Can you describe your transition from academia to a data scientist role?
CG: As a data scientist you get an opportunity to work with some amazingly rich data sets, and work on some extremely challenging problems. The work atmosphere is very research oriented, your team is almost entirely composed of PhDs, and in a lot of ways, it’s like working in academia, without having to spend large amounts of time fighting for grant money. This is what really attracted me to the Data Scientist role in the first place.
The transition from academia to industry was uneventful, though I realize that this is not the typical experience for most people. It helped that I had a very strong statistics background, and a very strong programming background, and I kind of hit the ground running once I was hired.
The process of getting hired was somewhat tedious because a lot of times companies screening resumes have no idea what to make of your resume when you present with a PhD in Mathematics, your resume points to your google scholar profile which lists paper after paper with the name E. Coli in the title, and you list a bunch of computer-sciency stuff in the skills section of your resume.
Increasingly, however, companies are getting better at identifying probable Data Scientists.
MP: What does your job involve?
CG: Many of the same things that PhDs love doing: talking to people about vaguely defined problems, working towards defining those problems to the point they can be solved, gathering or discovering data, gluing everything together, coming up with a solution, and then defending it in front of a lot of people.
Also, there’s usually pizza, cookies and donuts somewhere in there.
MP: What according to you are the key differences between working in academia and working in industry?
CG: I suppose every Data Scientist role is different, and the culture can be very different between companies. I find the two roles very similar, at least in the broad sense, in the sense of culture, working environment etc.
The main difference that I have felt is the pace (way faster). There are other minor superficial differences (I dress better now than I did while working in academia) which should not matter when considering the switch.
MP: What have you found most rewarding about your job? What have you found most frustrating?
CG: The most rewarding part of my job is coming up with solutions to problems or general insights. Sometimes you come up with a way of looking at the problem that no one has done before, and because you thought about it differently, your solution added a perspective that wasn’t available before. It’s incredibly rewarding to have that happen, and it feels great when you learn to make that happen frequently.
One thing PhDs are often bad at is understanding how organizations and individuals function, and as someone who never had to deal with that stuff before, it can sometimes be frustrating to have projects not move forward at the pace you want to push them. But it helps to realize that when you’re working in academia you’re kind of a one-man island (or a small team island), but working in a large corporation there could literally be hundreds of people who need to do tiny portions of things which enable the whole project. You may not even know how some things happen, or the people who make them happen, till they don’t, at which point you find out.
MP: What skills are needed to succeed as a data scientist?
CG: Be good at asking questions. Be good at understanding what you can conclude from a given dataset, and what you cannot. Develop the sense to understand what data can be had. Have the ability to handle pressure. Be creative. Always question your work; question the work of others.
I think these are skills most PhDs should already have.
MP: Do you have any suggestions for someone who wants to break into data science?
CG: I cannot stress this enough: learn to program, and be intellectually curious. Everything else will fall in place.
This work by ClubSciWri is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
