Status: 6YOE, Post-Grad from tier-1
Current Position: SMTS Salesforce
Location: Hyderabad
Date: June , 2024
I was interviewed for L62 level by Microsoft, posting this late but i was just caught up in the whole preparation/interviews.
I do have an offer and will be posting about that soon. :)
Round 1 : Online Coding Assessment (90 minutes)
Round 2 : Online In-Person coding round (1 hour)
The panel had an interviewer along with a shadow. After introductions we discussed the problem of finding needle in a haystack(String matching). Never in my 6 year career there was a need to revise KMP or Rabin-Karp hence i realised this is going to be a bumpy ride for the next one hour. Nevertheless i started with the vanilla implementation and stated what things might go wrong and how can we try to improve it using some prefix suffix precprocessing but interviewer was looking for a consolidated implementation which reduced the quadratic running time to linear.
After 45 minutes of awkward long gaps the interviewer decided to move on a design problem of creating an auth layer for some microservice. It was an extermely open-ended problem and i was not able to establish the expectations because of the limited time which again led to an extremely unstructured conversation.
At this point I didnt really think i would get another round but i did
Round 2: Round 2 : Online In-Person design round (1 hour)
Though this was supposed to be a design round the interviewer established we will be working on an algorithmic problem. This was most likely the result of the previous round. The interviewer asked a common problem of topological sort . For reference Leetcode problem of Course schedule II.
We discussed in greate detail how breadth/depth first traversals differ and how can they be used for solving this problem. I was able to complete the implementation. The interviewere was extremeley transparent with the feedback and extremely supportive throughout the process.
Verdict : [Reject]
Overall it was a good learning experience , it got me caliberated for scenarios where i do not have any idea about the question. My recommendation is to address the issue right from the start, rather than struggling with a challenge that required the combined efforts of three brilliant scientists to resolve.
Out of a meager experience of taking atleast 3 dozen of interviewes my advice to interviewers will be to try to ask questions which can be improved in an interative format and actually tests the problem solving.