Overall smooth interview process including combination of behavioral, coding, system design and research oriented questions. Through research oriented interviews you go through projects you have done and they ask questions about your work and then they propose an open problem and you should express your ideas. It is difficult to assess you performance on these interviews

1. Behavioral-style phone call 2. Simple data exercise screener 3. Virtual onsite. Several rounds, including a prepared presentation, and a 2-part data analysis exercise. I think I flubbed the SQL part of that. I was frustrated about the presentation though. The instructions said to take no more than an hour prepping it (ok lol) and to keep it VERY short, and NOT to go as far as, say, simulating data to chart. I felt like I bent the rules to fit in more info, ideas, analysis, how I expect the results to look - a bit like a grant proposal - and then I got dinged for not further breaking their own instructions and making it yet more in depth. Oh well, no one said this process has to be fair. So, word to the wise: ignore their instructions and make your deck way meatier! On the bright side I'm glad they gave feedback about which parts of the virtual onsite I flubbed. They were friendly and interesting to talk to. It mostly seemed like a process at least vaguely aligned with their hiring goals for the role, which is honestly more than I can say for most interview processes!

reached out by recruiter, first round is live coding interview in hackerrank with two questions, one on data transformation and the other is writing pseudo code to call preprocessing and a classification model object and calculate variance of performance metric