The first round of interview was setup with a Senior Applied Scientist and was focused on ML Depth and Breadth. The 60 mins of the interview was split into 2 parts - the first 30 mins were an elaborate technical discussion about my AI/ML projects and the next 30 mins was a question answer round about general ML topics. During the ML depth, the interviewer and I had a passionate discussion about my latest project which was in NLP. We started with laying out the problem statement and discussing the approaches that I took, what were my alternatives, the decision-making and the process and progress. Questions like what is the size of training data, how are you evaluating the performance of the model, what pre-processing techniques did you use, how do you plan to solve problem X were discussed. In the ML breadth part of the interview we started with basic discussions on loss functions, optimizers, linear regression, solving for overfitting, regularisation, feature selection and in depth discussion on transformers.

Applied for Amazon AGI. After first round, it will go into full round of multiple interviews. Lots of modern LLM training technic questions. There are still some behavioral questions, but less than general Amazon roles.