The process was shaped by an on-campus format with several elimination steps before anything resembling a final loop. It began with an OA that included frequency-based sorting, greedy and observation-based array work, an Amazon Leadership Principles work simulation, and a feedback and workstyle survey.
Round 1 was a 1-hour virtual interview using Zoom and a live coding platform. After a short introduction and a question about my preferred language, I was expected to start with a brute-force solution, then optimize it, dry-run it, and explain complexity. The two DSA questions involved shortest path in a grid with obstacles, including why BFS is better than DFS, avoiding cycles, and complexity. A modified minimum-walls-to-break version followed, with questions about the right data structure and whether nodes could be revisited.
Because I mentioned GenAI projects, we also discussed RAG hallucination verification and semantic search across huge logs with millions of lines, including chunking, embeddings, vector databases, and scaling. Round 2 leaned into behavioral and problem-solving evaluation, particularly leadership.
I was not offered the job. The difficult part was the compressed campus schedule joining leadership simulations, technical depth, RAG, and large-scale search with the coding work.
Location: Hyderābād. Overall feedback: Neutral. Offer status: No offer.
Discussion
Loading comments…