Skills to Succeed as an Applied Scientist

The July 2025 Sigecom Exchanges Newsletter has this really interesting article that our advisor reposted in our WeChat group:

EconCS in Industry: Skills to Succeed as an Applied Scientist

Devanur, Paes Leme and Schrijvers link

It applies more to EconCS students from the CS side, but the key points are universal for anyone wishing for industry I guess. Some key takeaways:

• Exploratory Data Analysis

  Exploratory Data Analysis is the iterative process of learning properties of the data that you’re working with. Typically, you start with a question or hypothesis (e.g. “the bids in an auction come from a Myerson-regular distribution”), then summarize, visualize, or model your data, and finally use what you learn to ask new questions.

  To me, this is really about developing the ability to understand real-world problems through data — and more importantly, to find new problems from the patterns you uncover. With LLMs making coding easier than ever, the harder skill now is building the logical framework to process data, extract structure, and abstract it into new ideas.

• Machine Learning and Statistics
  At minimum, take solid intro courses in both ML and Stats.

• Coding Mindset
  It’s not just about learning a language like Python or C++.
  (i) You need software-engineering-level intuition: how to read documentation, debug effectively, and review code thoughtfully.
  (ii) At a more advanced level, you should also be comfortable operating distributed systems.

One final thought: beyond technical skills, it really helps to stay open-minded. As students, we benefit from exploring ideas across disciplines and staying patient with unfamiliar jargon or frameworks — even when they seem unintuitive or “messy” at first. Some of the most interesting insights come from learning to see through someone else’s lens.