Meesho

Role: SDE 1 Lateral Hiring

College: Tier 1 college (2024 graduate)

Interview Date: 2024-10-19

Interview Type: Full-Time (Lateral Hiring)

Result: Rejected

Difficulty: Medium-Hard

Rounds: 3 rounds

Topics Asked: Data Structures, Algorithms, System Design, Heap, Concurrency, DBMS (Serializability), Project Discussion.

Experience:

The candidate applied for an SDE 1 lateral hiring role at Meesho through a referral. The candidate had 5 months of full-time and 2 months of internship experience.

Round 1: HackerRank OA (90 minutes)

This online assessment consisted of 3 questions with a medium-hard difficulty. The candidate successfully solved all questions.

One question involved an array and required using a divide operation (always by 2^i) to reach the minimum sum, with a solution approach using brute force and a priority queue (low-medium difficulty). Another question asked to implement Poll API/methods (create, delete, update, vote, print final results) with a rough idea of the poll object containing a list of questions. The candidate implemented all tasks except the bonus multi-threading part.

Round 2: DSA Interview

The interviewer expected complete and runnable code. The candidate was able to solve all the problems presented.

Specific questions included:

1. Minimum Number of Taps to Open to Water a Garden (LeetCode, hard).
2. Theoretical questions on the implementation of heap, heap sort, and heapify. The interviewer also asked if an element could be removed in log(n) time from a heap, to which the candidate suggested other data structures like a BST are better suited.

Round 3: System Design Interview

This round was entirely theoretical and started with basic HR questions and a discussion about projects at the candidate's current company.

Two design problems were given:

1. Implement a system where inventory is limited, and users need to be prevented from checking out if stock is too low (e.g., if only 5 items are left and 1000 people try to check out, 995 transactions should fail, and only 5 should succeed). The candidate discussed row locking and other concurrency methods, relating it to serializability from DBMS.
2. Design of game player allotment, where on a leaderboard, if a player joins, they need to be assigned a random player with almost the same rating, and then their rating needs to be updated. The candidate suggested an approach using hashing buckets.

The candidate felt they were not well-prepared for this round.