Microsoft | L62 | Aug 2025 [rejected]
Summary
I recently interviewed at Microsoft for an L62 position in August 2025 and unfortunately received a rejection. My interview process included a screening round, followed by four loop interviews covering DSA, System Design, and managerial aspects.
Full Experience
I applied for an L62 position at Microsoft through their online portal. The interview process began with a screening round.
During the screening round, the first 45 minutes were dedicated to a deep dive into my projects. Following that, for the next 30 minutes, I was presented with an open-ended graph question.
After successfully clearing the screening, I proceeded to the main loop interviews, which consisted of four distinct rounds:
- Round 1: Data Structures & Algorithms (DSA)
This round focused entirely on a specific LeetCode problem. - Round 2: System Design
I was tasked with designing an autocomplete search system, including trending search recommendations, similar to Google's functionality. - Round 3: System Design
This round involved designing an image upload and search system, akin to how Google fetches images based on text input. - Round 4: Managerial + DSA
This final round was a mix, starting with further project discussions, followed by several managerial questions, and concluding with another DSA problem.
Despite completing all rounds, I unfortunately received a rejection for the position.
Interview Questions (5)
Given an undirected graph, determine if it is a valid tree. A tree is characterized as a connected graph that contains no cycles.
Design a system that provides autocomplete suggestions as a user types, similar to Google's search bar. Additionally, the system should incorporate and display trending search recommendations.
Design a comprehensive system for uploading images and subsequently searching through them based on text input. The system should mimic the functionality of platforms like Google Images, where users can find relevant images by typing descriptive queries.
Given two nodes in a tree or a directed acyclic graph (DAG), where each node only contains information or a reference to its immediate parent (and no references to its children), find their lowest common ancestor (LCA).