Four rounds with the TT infra team.
- Given a graph where every node has a weight, pick a set of nodes such that no two chosen nodes are adjacent, and return the maximum sum.
- (The question for this round is just an unreadable placeholder in my original notes.)
- Given a graph, a source and a destination, return whether every path ends at the destination.
- Project deep dive plus an elementary-school math problem.
Every round also asked AI-related questions, for example how I use AI day to day to work more efficiently, and whether I had done anything to improve my team's efficiency.
All four rounds also threw a huge pile of fundamentals questions at me, including: K8s service IP vs cluster IP vs NodePort, walking through a request from the browser to the internal K8s cluster, DNS TTL vs negative TTL, gRPC vs HTTP, HTTP/1 vs HTTP/2 vs HTTP/3, how to debug a 5xx error, and K8s node vs pod.
All of the interviews were in English, but communication was a struggle. The team I interviewed with works on exactly this kind of stuff. Right after the interviews I got a verbal offer, and the next day they told me someone else had taken it.
Discussion
Loading comments…