My Amazon SDE-1 interview Experience - Rejected after Bar Raiser

I wanted to share my interview experience for the SDE-1 role at Amazon. The whole process had 5 rounds including the OA. OA Round I gave the OA around late October. It had 2 LeetCode medium questions.

I wanted to share my interview experience for the SDE-1 role at Amazon. The whole process had 5 rounds including the OA. OA Round I gave the OA around late October. It had 2 LeetCode medium questions. One was a sliding window problem. The other was based on number theory. If you were comfortable with prime numbers and sieve concepts, it was manageable. After solving the coding questions, there was also a 30-minute behavioral round in the OA. I didn’t hear anything for a while, but around late December I received an email saying they would like to move forward with interviews. Round 1 (Technical) They scheduled two technical rounds on the same day. In the first round, I was asked two DSA questions. Question 1: Course Schedule The interviewer asked me to implement both approaches: DFS (cycle detection) Kahn’s algorithm Then he asked a follow-up about reducing space complexity, since the goal was only to determine whether scheduling the courses is possible. Question 2: This was an interesting problem. You are given an array of prime numbers, and you generate a sequence using combinations of these primes.

Example: If primes = [2,3,5], the sequence becomes: 1, 2, 3, 4, 5, 6, 8, 9, 10, 12... The task was to find the nth number in this sequence. Initially it looked confusing, but after examining the test cases I realized it could be solved using a priority queue (min heap) approach. Once I got the idea, implementation was straightforward. I finished both questions in around 35 minutes. After that we discussed: my previous work some Leadership Principle questions what Amazon is currently working on Overall the round went very well. Round 2 (Technical) After a one hour break, I joined the second round feeling quite confident. The interviewer said he would ask two DSA questions. Question 1 You are given an array. Find the subarray whose (minimum + maximum) sum is maximum.

Example:

Input: [9,4,2,3,8,7] Answer: 15 Subarray: [8,7] All elements ai > 0. It turned out to be a greedy-style reasoning problem, and after dry running a few cases I figured it out. Then he asked a follow-up question What if we also want to return the maximum length subarray that achieves the same value? We discussed that variation briefly. Question 2 Distribute Coins in Binary Tree Luckily I had solved this problem before, so implementing it was straightforward. The round ended with a few more Leadership Principle questions. Round 3 (Technical + Discussion) The next week I had the third round. The interviewer introduced himself and explained the team and the type of work they do. He mentioned the round would include: one coding problem some discussion about Gen-AI The coding question was this one: https://carloarg02.medium.com/my-favorite-coding-question-to-give-candidates-17ea4758880c At first it looked simple, but the implementation was tricky. It took me around 30 minutes to arrive at a working solution. After that he asked whether I had worked on Gen-AI. I haven’t primarily worked in Gen-AI, but I explained some related experience and projects. Then we discussed my past projects and the round concluded. About one hour later, HR called and said I had cleared the round, and my Bar Raiser round would be scheduled the following week. Bar Raiser Round (Final) The Bar Raiser round unfortunately had some scheduling issues. One day before the interview, HR called and said it would be rescheduled to the next week. On the new date, I joined the meeting at the scheduled time, but no one joined. Later HR said the interviewer was waiting at a different time slot that I had not been informed about. Eventually it was rescheduled again for the next week. This round is where things went wrong. The interviewer focused heavily on Leadership Principles. At one point I discussed an example where we optimized an API and reduced latency from 10 seconds to 200 ms. I explained the architecture and the optimizations we applied. However, he kept pushing deeper into every intermediate step of the optimization process and wanted a very detailed breakdown of how the latency was reduced. I explained everything I knew, but I could sense the interviewer wasn’t fully convinced. The round lasted 1 hour 10 minutes, even though it was scheduled for 40 minutes. I spent most of that time explaining and answering follow-up questions. At that point I already had a feeling that I might have messed up. About a week later, I received the rejection email. Takeaway My main advice for anyone interviewing at Amazon: Prepare Leadership Principles thoroughly. Strong DSA performance is important, but LP answers can significantly affect your chances, especially in the Bar Raiser round.