The first interview was a call with the Engineering Manager. He was a very nice person. The interview consisted of general questions and discussions about machine learning. Some questions were about a use case actually done at Quora, and you had to think of an approach/pipeline for your machine learning solution, such as how to get data, feature engineering, which algorithm to use and why, etc.

After passing this interview, I got a coding test with four problems, similar to LeetCode. One problem was trivial, the second and third were medium-level, and the last one was difficult.

Which algorithms should be used in case of outliers?