Amazon DS new grad interview experience. Five rounds total — the first round was a screening, and the other four were split across two days, two rounds each day.
Of the five rounds, one was pure behavioral (BQ). The interviewer's title was Spec. Principal Sales Rep, and they asked "tell me about a time when you..." style questions the whole way through. Since the interviewer was a non-tech audience, whenever my story touched on something technical they'd follow up asking for the tech details — felt like they were testing how well I could explain technical details to a non-technical person.
The other three rounds (including the initial screening) were basically coding + stat/ml knowledge + resume project. Coding covered two types of questions:
- Pandas data manipulation: pretty easy, basically just groupby/agg/merge/concat gets it done.
- LC-style problems, but with no test cases — the focus was much more on explaining your thought process clearly than on getting a working solution. Of the three problems, the one closest to an actual LeetCode question was LC121; another was a graph problem solvable with BFS; and another felt similar to some Blind 75 problem, though I don't remember exactly which one. Overall not too hard — if you've gone through Blind 75 you can basically answer these.
The stat/ml knowledge questions were also fairly simple. I don't remember the specifics, but they'd basically throw out a topic and let you talk freely about it — the interviewer rarely interrupted, mostly waiting until I'd finished rambling before asking follow-ups. In one of these three rounds we also talked about the resume project, and they asked pretty detailed questions about the technical details of the project.
The last round was the hiring manager round. It was an ML design question about spam detection, again pretty open-ended in how you could answer it, and then they asked a bit of BQ on top of that.
Discussion
Loading comments…