You’re a seasoned data scientist. You keep up with the newest technologies and have proven your worth countless times. So, why does it feel like your career has stalled?
In order to level up, it might be worth taking another look at your non-technical skills, and how they might be improved. While tech prowess certainly powers the majority of data scientists’ success, strengthening areas like communication, writing and public speaking could be the missing pieces keeping techies from reaching their professional goals.
Because the advice to simply hone your soft skills can feel rather vague, Built In SF connected with Elliott Star, Chime’s director of data science. Star shared with us how he sharpened his non-technical abilities over the years, why that’s helped him excel in his role, and how other data scientists can follow suit.
Chime is a rapidly growing fintech that offers consumers access to checking and savings accounts without hidden or unnecessary fees. To help people gain more financial freedom, the company also offers tools to help build credit scores, receive their paychecks early and avoid overdraft fees.
What have been the most important non-technical skills in your career thus far?
First, excellence in written and verbal communication, perhaps even more than many of my technical and non-technical peers. As data scientists, it’s easy to forget that we are pretty deep in our data and machine learning domain. So even when we think we’re addressing items at a very cursory level, sometimes we just don’t communicate appropriately for our audience.
I also have the ability to handle ambiguity and improvise. Data science and machine learning share a ton of similarities with traditional software engineering teams, but we can’t expect to have the same level of predictability and operational consistency that software teams can sometimes rely on. We have to be able to roll with punches, have backup plans on top of backup plans and work outside of our comfort zone. That’s just a part of our career.
Finally, I display excellence in basic execution. Especially during Covid-19, when folks are predominantly working remotely, doing the little things really well keeps everything running smoothly. That means efficiency in emails, keeping Slack messages succinct, clear timing on meetings, effective calendar management and scheduling, and being able to share screens quickly and without friction. The minutes shaved off here and there may not seem like a lot, but they add up.
The best data scientists know when not to do data science.”
Looking back, what are some of the professional or personal experiences that have helped you develop those non-technical skills?
I did a lot of public speaking growing up, which helped me feel really confident presenting and conveying my ideas and thoughts to widely varying crowds. Similarly, I had a mixed background from an educational standpoint: At one point, I was considering a double major in physics and political science, so I got a lot of repetitions in with short and long-form writing. Obviously, it’s hard to go back in time and change what you studied, but even now in your career, getting lots of repetitions definitely helps, especially when you can get really solid feedback on your writing.
Secondly, I worked on projects with a high probability of failure; or ones where the technical aspects felt comfortable, but all the associated non-tech aspects felt very uncomfortable. At the time, I would not describe these as enjoyable experiences by any stretch of the imagination, but they were definitely the ones that really forced me to advance in my non-technical skillset.
Beyond building and maintaining robust analytics and coding skills, what advice would you give to a talented data scientist wanting to grow their career?
Get really good at the fundamentals of logical reasoning and analysis. I see a surprising number of aspiring data scientists and ML engineers who miss layups when it comes to basic analysis and logical thinking. One of the greatest data scientists I ever worked with once told me that “the best data scientists know when not to do data science.” That really stuck with me as being a key differentiator, especially as tools and automation continue to advance our field.
At the same time, learn how to adequately express your limits. The last thing you want to do is back yourself into a corner by advertising you have a specific skill set and then be asked to deliver within that. It’s OK to say that you aren’t familiar with something: B.S.-ing your way through situations may work in the short term, but will always come back around to bite you.
