Three interviews were conducted in a back-to-back format, referred to as "the loop." They asked both LPs (behavioral questions) and technical questions, and these were split fairly evenly within each interview. This process was for an AI/ML Software Engineering position.

The first interview's technical question was about an LRU cache. The second involved a DFS on packet launches. The third was a really weird bash script task to return the locations of certain files based on their suffix. Wasn't incredibly difficult, but the questions were stated so strangely and were so out of place that I was caught off guard.