Walk me through one of your projects. How did you monitor? How did you detect drift or bias? Explain a time to me when a model failed. How did you detect the failure? There was also a coding interview in which they shared a GitHub project beforehand. Task was to then pair-code with one of their engineers.
- Asked some basic ML questions like what is bias and variance - How do you train a NN network
Was mostly with Pandas and distributed computing
Implement convolution operation from scratch in python
They ask me about tech skills related to deep learning and if I had experience building platforms to run different DL models on various types of processors
Work through the resume and dive very deep into the skills listed in the resume. Question for the project and the position. A question about a specific question about one challenge in their current pipeline, how to tackle them
System design questions with different theoretical questions mixed in
Describe a recent research paper you found interesting.
2 algorithmic interviews and 2 ML interviews. For algorithmic, both were easy to medium leetcode-style questions. For ML, one was about basics of ML, statistics, data analysis and the other was little more specific topic like NLP.
How do you keep up with the latest research?
Viewing 6631 - 6640 interview questions