Initially, there was a take-home coding test, followed by four interviews in one day. One interview focused on general ML knowledge, the second on Python-related questions, the third on PyTorch coding questions, and the fourth on behavioral aspects.

recruiter → OA → Four 50-min in-person interviews (back-to-back, 1 behavioral, 1 role specific, 1 general technical, 1 hands-on). OA was easy. role specific asks ML fundamentals and concepts, general asks C++ and Python concepts