Amazon SDE II Software Development Engineer Interview Experience
Amazon · Mid level · Software Engineer
I went in expecting the usual behavioral-first format, but in my loop they flipped it and opened with the coding question. The behavioral part still mattered, they just used the follow-ups to see if I actually knew my own work in detail.
Interview process
I interviewed for an SDE II role, and the whole process felt pretty standard and transparent. Amazon told me upfront what the process would look like, so I did not feel like I was guessing what came next: first an online assessment, then three 40-minute interviews back to back on the same day. The technical questions were all very classic problems, and the biggest surprise was that my interviewers often started with coding first instead of spending the first part on behavioral. The behavioral side was still important, but the follow-ups went deep into the actual details of my work instead of stopping at a polished STAR answer. After the loop, my recruiter mostly asked about logistics like start date and work authorization, and then they sent me the offer letter.
Interview rounds · 4
- 1
Online assessment
CodingData Structures & AlgorithmsBehavioralI started with an online assessment that had two coding questions and then a separate leadership-principles/personality section, and they had already told me the structure upfront so nothing about it felt surprising.
- Q1. Determine the minimum starting health for a player to win the game, where each round reduces health by power[i], health must stay above zero, and armor can block min(armor, power[i]) in one round.
- Q2. Given an array of daily stock returns and at most k deletions, determine the maximum number of contiguous days with the same return value after omitting up to k days.
Q3. In the leadership-principles section, I got agree/disagree statements around scenarios like choosing between two features, helping a teammate after hours, reporting conflict, and general work-style statements.
How they answeredIt felt more like LP and personality alignment than open-ended behavioral questions. I remember the format was strongly disagree to strongly agree, and none of the statements felt especially tricky to me.
- 2
Final / onsite round
CodingData Structures & AlgorithmsBehavioralMy first live round was different from what I had read online because they started with coding first and then used the remaining time for behavioral, and the whole thing felt more like a detailed chat than a stress interview.
Q1. Implement regular expression matching for a string and pattern, where . matches any single character and * matches zero or more of the preceding element.
How they answeredI implemented the standard regex-matching version with dot and star and talked through my logic as I coded it. After that, the interviewer added a bracket follow-up, but he did not ask me to implement it, so I just explained out loud how my thinking would change and how I would handle parsing that case.
Follow-up questions- What if the pattern also included opening and closing brackets? How would your solution change?
Q2. Tell me about a time you had to make a decision with incomplete information. What did you do?
How they answeredI answered with a story from a project I was working on around regression testing and used the STAR format. What stood out was that they kept drilling into the details like the framework, how often the regression ran, and what I personally owned. It did not feel aggressive to me. It felt more like a conversation to check that I really knew the work deeply, not just the headline story.
Follow-up questions- What platform or framework were you using to run that regression?
- How frequently were you running the regression?
- What exactly was your responsibility in that work?
- 3
Final / onsite round
CodingData Structures & AlgorithmsBehavioralThe second round felt more collaborative because the coding problem was a very standard one and this interviewer gave me a lot of hints and seemed pretty satisfied with my answer.
Q1. Given a string of digits from 2 to 9 and the classic phone keypad mapping, return all possible letter combinations the number could represent.
How they answeredI recognized it as the usual phone-number letter combinations problem and worked through it from the keypad mapping to the full set of combinations. This interviewer was pretty flexible and gave me a lot of hints, so the round felt collaborative rather than rigid.
Q2. Talk about a project or a time when you took ownership. What were your responsibilities and what actions did you take?
How they answeredI used a project where I had taken clear ownership and explained what I was responsible for and the actions I drove myself. The follow-up was the useful part because they were not just looking for a success story. They wanted to hear whether I had regrets and what I would change if I could do it again.
Follow-up questions- Did you have any regrets?
- If you could go back, would you do it differently?
- 4
Final / onsite round
CodingData Structures & AlgorithmsBehavioralThe last round was another classic coding question, and the behavioral part again focused on disagreement, judgment, and what I learned afterward rather than just whether I fixed the problem.
Q1. Given a number of courses and a list of prerequisite pairs, return true or false depending on whether it is possible to finish all courses.
How they answeredI treated it as the classic course-schedule problem and explained how to decide whether all the courses could be completed given the prerequisite dependencies.
Q2. Tell me about a time you disagreed with a teammate and how you handled that conflict.
How they answeredI answered with a disagreement story and walked through how I handled the conflict. Then they pushed on reflection and asked what else I could have done better, so it was not enough just to say I resolved it. They wanted to hear what I learned and what I would improve next time.
Follow-up questions- Was there any other action you should have taken?
- What else do you think you could have done better in that situation?
Tips from the candidate
I would definitely prep the classic Amazon-style coding questions because mine were all recognizable patterns, not weird trick questions. The bigger thing is do not just memorize STAR and think that is enough. You need to know the real details behind every story you plan to use, like the framework, how often something ran, what exactly you owned, and what you would do differently, because that is where they keep digging. I would also expect the OA behavioral part to be LP-style agree/disagree statements, not open-ended answers.
Company culture
It felt very process-driven and transparent to me. They were clear upfront about the structure, and even the OA behavioral section was obviously tied to leadership principles. The interviewers had different styles, but even when they drilled into my answers it did not feel like they were trying to intimidate me. It felt more like they wanted proof that I actually owned the work and could talk about it in detail. Once the loop was done, the recruiting side moved pretty quickly into logistics and then the offer.