SWE intern
Interview Date
22-10-2024
Result
Rejected
Difficulty
Medium
Rounds
03
Drive Type
Off-Campus
Topics asked
Detailed experience
I applied for the Microsoft SWE Intern role through the off-campus process and was first invited for an online coding assessment. The OA was conducted on Mettl and had three coding problems. The first was an array-based problem and was relatively easy, the second involved trees and was around easy-medium level, while the third was a medium-level DP problem. I managed to solve all three, which helped me move forward. The next stage was the Group-Fly round. This was different from a normal online interview because it was a pen-and-paper coding round. We were given two problems and had only around 30 minutes. We had to write complete working code, including input/output, and also mention the time and space complexity along with alternative approaches. The first problem was based on bit manipulation, while the second involved checking whether a given word could be formed by combining one or more words from a given list. The time constraint made this round quite intense because there was little room for trial and error. After that, the interviewer discussed an ML-based problem and asked me to explain my approach. I was then given a smaller ad-hoc coding problem, which I managed to solve fairly quickly. The round ended with some behavioral questions, including how I would handle a situation where I was unable to meet my manager's expectations. I cleared this round and moved to the final technical/project discussion. The final round was comparatively more conversational. The interviewer initially asked about my interests outside academics and then moved to my projects. I had worked on both a web-development project and an ML project, and the interviewer asked if he could look at my GitHub repository. Most of the discussion focused on my ML project and the reasoning behind different implementation choices. I was asked about concepts such as feature scaling, label encoding, curse of dimensionality, clustering techniques, and Bayesian classifiers. The interviewer kept going deeper into the project instead of simply asking textbook definitions, so I had to explain why I had chosen particular approaches. We also discussed my future goals and what I wanted to work on as a software engineer. The final interview lasted around 40 minutes. After the interviews, I had to wait for the final decision. About a month later, I received the confirmation that I had been selected for the Microsoft internship. Overall, the process was not limited to DSA. The coding rounds tested problem-solving speed, while the final discussion heavily depended on how well I actually understood the projects mentioned on my resume.