Round 1 — Technical Screening
A mix of aptitude and coding:
- Aptitude + DI: Data interpretation, quantitative aptitude, logical/problem-solving, unseen passages/comprehension
- Coding: Fibonacci series and basic mathematical/algorithmic problems
Time management mattered — many different question types. The coding was easier than later rounds, but you need to convert math ideas to code quickly.
Prep: Arrays, strings, loops, recursion, math patterns, Fibonacci/series, aptitude — under time pressure.
Round 2 — Deep Technical Interview
The interviewer roamed across projects, DSA, databases, OOP, system design, and puzzles. They can pick anything on your resume and drill down — know your projects cold.
Projects + Tech Stack: For every technology, be ready to explain: why you used it, what problem it solves, alternatives, pros/cons, and what happens internally. For databases: SQL vs NoSQL, indexing, normalization, transactions, caching, Redis, scalability.
DSA problems:
- Number of Islands (grid/DFS/BFS/DSU)
- Search in Rotated Sorted Array (modified binary search, O(log n))
- Disjoint Set Union (path compression, union by rank)
- LRU Cache (HashMap + Doubly Linked List)
Puzzle: Bridge and Torch — 4 people (1, 2, 7, 10 min), one torch, max 2 cross at a time, slower person's time counts. Minimum time = 17 minutes (send the two fastest as ferries strategically).
Production Troubleshooting: "Your server suddenly became slow this morning. What do you do?" — they want a structured approach: confirm the problem (CPU/memory/latency/errors), check what changed (deployments, config, traffic), find bottlenecks (slow queries, N+1, memory leaks), check infra, then mitigate first (rollback/scale) before deep investigation.
Handling a 50GB file: Don't load it all into memory — stream/chunk it. Discuss buffered I/O, parallel processing, object storage (S3), message queues.
OOP: Singleton pattern (private constructor, static instance, global access method).
Round 3 — Design Coding + Project
Tab Handler Design: Design a browser tab manager — open/close tabs, unique IDs, find current tab, list all. Expected: HashMap (ID → Node) + Doubly Linked List for order and O(1) operations.
Middleware: What is it, why used, request/response lifecycle, auth/validation/error-handling middleware, how it differs from a controller.
Rate Limiter design: Fixed window, sliding window, sliding window log, token bucket, leaky bucket. Plus Redis for distributed limiting, race conditions, TTL, per-user vs per-IP.
Caching: Cache hit/miss, Redis, cache-aside, write-through, write-back, TTL, invalidation, eviction (LRU), what happens when cache goes down. Follow-up: "Why not store everything in Redis?" (cost, memory, persistence, consistency).
C++ vs Java speed: Nuanced answer expected — not just "compiled vs interpreted." C++ compiles to native code, Java to bytecode + JIT (which can be very fast), C++ has more memory control, Java has GC overhead. Performance depends on workload.
Round 4 — More Technical
DSA questions, project deep-dive, lots on APIs in the personal project, and SQL queries.
One coding question: Second rightmost element in a binary tree — clarify what "rightmost" means, then use level-order traversal (BFS).
Round 5 — HR
"Why do you want to join InfoEdge" and standard HR questions. Operating System questions were also asked.
Tips for InfoEdge
- Know your resume inside out — they drill into every technology and project. Prepare the "why/alternatives/internals" for each.
- CS fundamentals are heavily tested — DBMS, OOP, OS, caching, system design basics.
- DSA covers standard patterns — islands/grids, binary search, DSU, LRU, trees.
- Practice puzzles — Bridge and Torch and similar classics.
- Prepare production/systems thinking — "server is slow," "handle a huge file" — structured debugging answers.
- GfG previous-year questions help for InfoEdge specifically.