A recruiter reached out to me.
Phone screen: DFS, pretty standard.
I had the next four rounds on the same day:
Round 1: ML fundamentals. Pure conversation, talked about my projects and experience, and he'd find things to ask about along the way. Felt pretty smooth, the guy listened carefully and didn't interrupt randomly, very respectful.
Round 2: ML coding. Based on the prep materials they gave me, fairly easy.
Round 3: ML system design. This round was a mess, he controlled the whole pace. He started by asking about my experience, spent fifteen minutes on that, then said it was taking too long. Once we got into the ML design itself, he kept interrupting, redirecting the conversation, then complaining there were too many details. Kind of confused about how people are supposed to handle that...
Round 4: DSA. This round was ridiculous, an interviewer who used to work at Google gave me a problem I'd never seen before. There's a 2D maze with an exit E, and the other cells are either walls or empty. There's a robot that can move up, down, left, or right, but the robot doesn't know the layout of the maze. When the robot hits a wall, it just stays where it is. The question: how do you generate a single sequence of instructions (some combination of up/down/left/right moves) such that the robot, no matter where it starts in the maze, will reach exit E as long as it follows that sequence. Assume E is never blocked off, and that at least one solution exists. I didn't write a single line of code. Never seen this type of problem before. It took me two days after the interview to actually figure out the solution, and once I saw it, it turned out to be simple — but during the interview, no matter what hints he gave me, I couldn't get there.
Two weeks of prep, and I got tripped up by one DSA question. Feeling pretty rough about it, and not sure how to improve from here. Looking for guidance.
Discussion
Loading comments…