Three Months Into Industry

6 minute read

Published:

I “left” (I say left in quotations because I’m still working on my Ph.D part time) approximately 3 months ago. That is a bit of a milestone. It is a quarter of a year working at a senior-ish level as a data scientist at a national bank. I know a lot of PhDs, especially in quantitative disciplines, are thinking of making the jump from academia to industry. This sequence of blog posts is not advice on how to make that jump, but rather to document one perspective on what that change entails.

I’m not sure how best to document my experiences, but the most interesting comparisons to me are:

  • Differences in Challenges: How does working on a PhD differ from working in industry with respect to the challenges you face? Are obstacles easier/harder to overcome? What are the differences in obstacles?

  • Differences in Work Satisfaction: How do I like working on a PhD versus working in industry? Is the work better/worse? More/less interesting?

  • Differences in Work Life Balance: This one is pretty self explanatory.

Let’s begin.

Differences in Challenges

One of the earliest challenges I had to face was to find stuff to do! I imagine this is not typical of most data science positions. You’re usually hired into a data science team, who is likely fostering a bunch of projects. Not me, I was hired into an app team comprised almost entirely of developers in Java and kdb+ (or are they q developers? It doesn’t matter). The team is new, I joined less than a year after the team was formed. Consequently, most of the work being done is development work. It was soon revealed to me that (I’m paraphrasing here) the team did not want a data scientist, they wanted another developer. But the bank works in mysterious bureaucratic ways and so they were handed a data scientist to “inject AI into trade compliance”. However, because the team is so new and there is much development work to be done, machine learning and data science just isn’t on their radar right now. So my first challenge was to find something to do, or more precisely, to understand the processes of the business we support and find areas where I could add value with machine learning.

That’s tough. It requires a skill I did not properly hone in my time in academia: talking to people and empathizing. It would be easy for me to just create a model to do give analysts another number to look at (or ignore, as a manager has said to me), but the trick to being a good data scientist is creating solutions which address an actual need and not solve some sort of math problem. You can see where I (a person who has spent a lot of his life just solving math problems) could encounter some difficulty.

Differences in Work Satisfaction

I said this to my manager fairly early on: “I don’t care much about financial markets”. I’m a boring investor. I put a set amount of money into an index fund every month automatically. I don’t care about Gamestop (save for the memes), I could give a shit about SPY and TSLA calls, and those are the interesting bits! I work in fixed income! If I find Telsa drama boring, imagine how I feel about government bonds!

Suffice to say, my work at the hospital was orders of magnitude more interesting and fulfilling (though the pay here is much at the bank. I will note here that there is an interesting inverse relationship between how fulfilling a job is and the pay for those jobs. More on that in Bullshit Jobs, an excellent book in which data scientists are called out by name. I digress… ). That being said, there are other ways I get work satisfaction.

Though I don’t (yet) have access to enormous compute I do have access to enormous data. Consequently, the problems become interesting when I can abstract away the boring financial details and focus on the underlying statistical or math problem. That is an enormous difference between grad school and industry. I finally have the data I need to (at least approximately) answer the questions I want to answer.

Differences in Work Life Balance

Unsurprisingly, work life balance is better in industry. In grad school, the guilt to work is a long, dull, pain. There is always work to be done, and if you aren’t doing it then (you tell yourself) you are lazy. In industry, the guilt to work is a sharp quick pain that comes in waves. There is always work to be done, but I know I’m going to come back to it tomorrow for the same amount of time so I don’t feel as bad about taking a 40 minute break, or working on something else.

That being said, there are times I’ve had to work after 5 pm. That’s fine with me, its pretty infrequent, but I have been witness to people scheduling meetings at 8 pm as if I didn’t have a life outside work. And even stranger, I’ve seen people decline those meetings because they are already in meetings at 8pm. What the fuck? What the actual fuck?

I’m not interested in imputing why people do this. I benefit from low amounts of responsibility (no mortgage, no kids, no family) and so perhaps I’m not as motivated by promotions, bonuses, or being fired as some people are. That will maybe change as I grow a bit older, but I’d sooner quit than schedule a meeting that late into the night. Famous last words.

I’ll revisit this story in another 3 months, which would be around November. In the meantime, if you have additional questions or would like to chat about my experience (or yours for that matter, I’d much rather listen than talk), please reach out via twitter.