Machine Learning Engineer Interview Experiences

Couldn't recall the DBMS details and that cost me. OA was fine. Technical asked about indexing and race conditions, stuff I knew but hadn't reviewed. Then had to explain an ML project end to end including loss function choice.

Their ML OA is completely different from SWE, 3 Python questions focused on linear algebra and NumPy. then the technical interview was ML system design and they went into questions about residual connections and in context learning in LLMs. you need solid deep learning knowledge not just leetcode skills

Was expecting heavy technicals going into the first round but really just SQL queries and theory about how you would go about implementing some APIs for financial data was surprised there was no live coding.

HireVue first, 6 questions 30 sec prep for each, favorite data structure came up then HackerRank had two medium and hard, solved one fully and partially the other and never got to superday.

Single hour call with two people about past ML projects. Discussed models I'd built, data I used and how I evaluated them. Made it to the assignment stage but no offer after submitting

They liked anything location data or time series related given what Ola does. Aptitude was fine. Technical interviews covered ML projects then coding on arrays and graphs. HR had a guesstimate on driver supply.