One interviewer end my interview 20 mins earlier than expected and he failed me. But other interviewer think I did pretty good. So they downgrade my level. I could not get how the interviewer got the conclusion that I'm not good enough for the role within only 40 mins? If they can decide in 40 mins, then why the interview process designed as 60 mins? It's unfair and ridiculous!
Surprisingly easy — I expected tougher questions, but the coding round felt more like a warm-up. The main challenge was a DSA problem about counting islands in a 2D grid, which led to a discussion on DFS versus BFS and handling large grids. Funny enough, I had revisited that exact type of question while prepping on PracHub, which made me feel more confident. The interview wrapped up with a behavioral round, and I accepted an offer, but ultimately decided to decline it for another opportunity. Overall, it was a smooth experience.
Interview questions [1]
Question 1
Number of Islands — given a 2D grid of '1's (land) and '0's (water), count the number of connected islands. Walk through DFS vs BFS, and discuss how to avoid revisiting cells (in-place mutation vs visited set) and what changes if the grid is huge and must stream from disk.
It started with an OA, and then after a few weeks, I got invited to four rounds of interviews: technical and behavioral at 3 of the 4, and behavioral only at one.
I applied online. I interviewed at Amazon (Calgary, AB) in Jun 2026
Interview
Online Assessment is the first step in the process. I didn’t have an HR phone screening and went straight to the OA after applying. It was sent to me about a week after I submitted my application.
Interview questions [1]
Question 1
The first question is LeetCode style algorithms question, and the second question gives a full stack repo (choice of Java, NodeJS, or Django) and asks to solve a backend issue which is causing a bug in the frontend. Unit tests must pass to pass the second question. You can run both backend/frontend indivdually or together