The application process consists of a technical screening round, followed by an onsite round if you pass. The onsite round includes six stages: one tech talk and five technical interviews, including ML depth, ML breadth, ML application, and two coding interviews. While each interview is manageable on its own and there is a wealth of online resources available for preparation, the overall process can feel overwhelming due to its intensity and scope. I also found certain aspects of the procedure to be challenging: - Frequent changes to information: There were instances where details about the process or expectations were updated unexpectedly, making preparation more difficult. - Timing issues: Some interviews started later than scheduled, and many extended beyond their planned durations, which disrupted the overall schedule. - Irrelevant questions: A few interview questions seemed unrelated to the role, making it unclear how they aligned with the position’s requirements. One key note for potential candidates: Amazon places a significant emphasis on behavioral questions, which can constitute up to 50% of the interview process. It's essential to be well-prepared for these as they are a critical part of the evaluation.

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.