I attended three interviews. The first was about the fundamentals of ML, DL, and NLP. The second round is an AI/ML coding round, which is some kind of open-ended coding questions, and the final round is a project discussion with questions about transformers and other NLP topics. The interview structure is different, but neither good nor bad, which can be difficult for candidates to learn from. If they do not respond within 2-3 workdays, you are rejected.

Technical round and HR round both are well-focused on DSA questions. C, C++, Python, Java and mostly Graphs, Trees and other data structures. In java object oriented programming and collections topic is important