Overview of the Data Science Interview Process

Talk by Michelle Gasbon and Steve Greenberg from Kaggle CareerCon 2018 including my summary notes.

Prepare to Prepare

Being interviewed for a job is a skill, and can thus be trained. Make more time than you think you’ll need before interviews and prepare.

Make a prioritised list of opportunities, first go through the top interesting opportunities before starting process with less interesting ones. This is so you don’t get started in a more interesting opportunity while being in the end of another process. Further, you only have bandwidth for x processes to run simultaneously if you want to be able to prepare properly.

Actual Preparation

Research the company well - reach out to people you might know that works/have worked at the company. Know the company’s core values. Know who you’ll be talking to in each interview.

Practice out loud - preferably in front of an audience. Get feedback. Alternatively, record yourself and watch.

Structure your answer - Before going in to details answering a question, start with a high level summary. This might be enough for the interviewer who then can choose to move on to another topic.

Specific Knowledge

Make sure your resume covers your knowledge areas that are relevant but don’t overstate them.

If you don’t have the answer, then you might want to cover how you would go about learning the answer.

Coding challenges

  • Focus is on seeing how you think about the problem
  • Make sure you understand the problem before you start

Math & Stats

  • How would you model a situation? Maybe two different ways and pros and cons. Then how to implement them.
  • “Tell me about an insight you’ve found and its impact.”

Soft skills

  • Prepare a list of anecdotes that gives an idea of how you like to work.
  • Not only about successes, but only about what you learned along the way.
  • How would you contribute to team culture?
  • Be intellectually humble.

Comments