I was contacted by a technical recruiter, followed by an amazing conversation on the phone about my technical interests and experience, which led to a phone screen.
The phone screen was interesting. It started with a behavioral phase where I had to explain a little about myself and an interesting project I completed. The interviewer was clearly not happy with the project I described.
This was followed by a series of Machine Learning questions, such as what technologies (classification, etc.) I have worked on and what algorithms I used previously. Next was about overfitting: how I identify and overcome it. The interviewer was unhappy with the usual bias vs. variance, cross-validation, re-sampling, feature engineering, regularization, and other methods (ensemble, etc.)-based answers and had no intention of giving a hint of what was being expected. There were also some routine machine learning questions that tested basic ML knowledge.
This was followed by a coding challenge on Codepair. It was an easy problem straight from LeetCode (string + DP), but the funny part was the interviewer was pissed off at the DP solution and demanded a 2^n recursive solution. At this point, it was clear to me that the interview had derailed a long time back, and it felt like the interviewer was not at all interested in making the call.
The obvious rejection came a week later on a holiday in the usual email format.
PS: The interviewer constantly yelled whenever he did not like my answer and made some unprofessional remarks about how people lack experience in dealing with problems.
String-based search problem.
(Easy difficulty on LeetCode, direct)
The following metrics were computed from 8 interview experiences for the Zillow Machine Learning Engineer role in United States.
Zillow's interview process for their Machine Learning Engineer roles in the United States is extremely selective, failing the vast majority of engineers.
Candidates reported having very negative feelings for Zillow's Machine Learning Engineer interview process in United States.