Microsoft Data Scientist interview experience

I had three rounds back to back, and most of it was behavioral questions paired with a deep dive into my projects. Instead of writing much code, they pressed me hard on the logic behind every decision I made—why I chose a particular approach, what tradeoffs I accepted, and how I interpreted results.

One of the technical-focused moments centered on data science fundamentals around data quality, especially outliers. I was asked what I’d do if a dataset had feature values that were extremely large—whether I’d remove them, keep them, or transform them—and then I got a follow-up that probed