Four interviews:

• 15m call with in-house recruiter about role.
• 1h no coding problem solving interview. I was given a deep learning case study and had to talk about how I'd go about solving it.
• 1h programming interview. Data structures and algorithms.
• 1h PyTorch debugging interview. I was given a Google colab notebook which trained a CNN, and I had to debug errors.

Given a dataset containing telemetry and video data from autonomous vehicles, how would you pull out all examples when the vehicle needed to make an emergency brake due to being cut off by another vehicle on the road?