Amazon Intern interview Experience | 31 October 2026 | DSA + ML + GenAI

Anonymous User
2026
Nov 01, 2026
Nov 11, 2026

Amazon Interview Experience | 31 October 2026 | DSA + ML + GenAI

Education: B.Tech in CSE (Tier-3 College)
Experience: Data Science Intern at an AI Startup


Online Assessment (OA)

Applied through the job portal and received the OA link on the same day.

The OA consisted of:

  • 1 DSA question
  • Code debugging
  • Behavioural questions

After a month, I received an email inviting me to the hiring drive. I opted for offline interviews, as I had heard that candidates who chose online interviews were often rejected without being scheduled for further rounds.

Venue: IGDTUW, Delhi
Date: 31/10/2026


1st Round: 3 DSA Questions + Behavioural

The interviewer introduced himself and then asked me to give my introduction. After that, he handed me a pen and paper to solve the questions.

1. Find the maximum sum among the longest paths in a binary tree.

Similar to the basic path sum problem, but with an additional state to track the path length.

I took a minute to understand the question and then explained two possible approaches:

  1. Two-pass approach: First calculate the maximum path length and then find the maximum sum among paths of that length.
  2. Single-pass approach: Use DFS while maintaining both the sum and length as states.

The interviewer pointed out an edge case in my approach, which I corrected. He then asked me to write the code on paper, dry-run it, and explain its time and space complexities.

2. Group all packages together using minimum swaps.

Similar to Minimum Swaps to Group All 1s Together II.

I had solved a similar problem before, so I started by explaining the brute-force approach. I then mentioned that it would not be efficient for larger constraints and optimized it using a sliding window.

I wrote the optimized code, performed a dry run, and explained its time and space complexities.

3. LCA.

Pretty state forward LCA question. I implemented the recursive solution. He asked me if I knew any optimizations on this , so I mentioned Binary Lifting. He asked if I could code it up , so I did . We did a few dry runs and everything worked.

Behavioural Questions

  • Tell me about a time when you did something outside your comfort zone.
  • Tell me about a time when someone criticized your decision. How did you react, and how did you handle the situation?

I answered both questions using the STAR method and incorporated relevant Amazon Leadership Principles.

Result: Cleared the first round! Just five minutes later, my second round was scheduled.


2nd Round: Probability Puzzle + DSA + ML + GenAI

Honestly, I went into the room expecting an HR round, but it turned out to be another technical interview.

1. Probability Puzzle

The interviewer introduced himself and asked me to introduce myself as well. He then presented a probability puzzle.

This was probably the only part of the interview where I struggled. I wasn't particularly prepared for probability-based questions.

We discussed the problem for around five minutes. I was gradually approaching a solution, and the interviewer also helped me derive the final approach.

2. DSA: Currency Exchange Calculator

We had to implement a currency exchange calculator.

The input consisted of three fields:

[currency1, currency2, exchangeRate]

We were also given queries in the following format:

[currency1, currency2, amountInCurrency1]

I approached the problem using a graph, where each currency represented a node and the exchange rates represented weighted edges.

I suggested creating a directed graph and adding reverse edges with the inverse exchange rates to handle conversions in both directions.

The interviewer asked me to implement the solution, followed by a dry run.

I made a small mistake by initially assuming that the graph wouldn't contain cycles. As a follow-up, he asked how I would handle cycles, so I incorporated a visited state to avoid revisiting nodes during traversal.

3. Machine Learning and Generative AI

The interviewer asked me about:

  • Where I use LLMs and how they have helped me.
  • How CNNs work and their underlying architecture.
  • How Transformers work and their underlying architecture.
  • The role of encoders and decoders in both CNNs and Transformers.

He also asked me to draw diagrams explaining the architectures of CNNs and Transformers.


Final Feedback

At the end of the interview, I asked the interviewer how I had performed and whether there were any areas I could improve.

His feedback was:

"You're a strong candidate. You should work on defending everything you have mentioned on your resume. Since you've mentioned that you've been working with Machine Learning, you should also have a good understanding of probability and statistics."


STATUS : GOT SELECTEDDDDDDDDD

Comments (12)