Very first step was a technical screen. Was directed to a shared coding site. Asked to write a couple of SQL queries that were moderately difficult. However, because this is the first step, they were more interested in the thought process and didn’t worry too much about optimizing. Then there were Python coding questions. They allowed you to Google syntax and whatever you needed. Mostly working with pandas data frames. Last question was a short business case testing business sense and some stats, mostly product-focused (what metrics would you use/calculate, what would you check if numbers seem wrong, etc). Next round was an interview with the hiring manager, with basic introductions and more “what metrics” and business sense questions, walkthrough of resume, tell me about a project, etc. Final round was 4.5 hours long, panel interview. One was a continuation of the SQL/Python coding, this time with attention to optimizing and efficiency, and generally harder questions. Another interview was on A/B testing, stats, and design of experiments. One was business sense and metrics, a couple resume walkthroughs with some stats/ML technical questions asked (what algorithm would you use and why, how would you check whether your model worked), and finally a conversation with HR. In the end, I wasn’t chosen. It was a very lengthy process, but overall a positive one. The interviewers were never out to “stump” me and very happily helped me when stuck. They want you to succeed.