← Back to Directory

Amazon interview notes, page 136 of 159

FAANG / Big Tech

Most of these Amazon write-ups come from Software Engineers (1,092), then Data Scientists (182) and Product Managers (165). The most reported rounds are the technical phone screen (618), the behavioral / leadership round (309), the online assessment (228) and the system design round (159).

1581 interview notes · updated Oct 2026

Amazon·Technical Program Manager·Onsite - Behavioral / Leadership

Jun 2026
TPM loop at Amazon, behavioral focus with at least one question digging into how you actually managed headcount and budget on a real project. Pretty standard for the level but they want specifics.
  • Can you walk through a time you were responsible for managing headcount and budget on a project you led?

“I had a decent example but fumbled the numbers part.”

View Post

Amazon·Software Engineer·Technical Phone Screen

Jun 2026
Amazon SWE interview with an AI-assisted debugging round focused on a return-eligibility service that was incorrectly approving high-risk return requests. The problem was interesting but the breadth of what they expected was a lot to cover in one session.
  • You're working on an e-commerce return-eligibility service that uses a risk score to decide whether items can be returned. The system has recently been approving returns it shouldn't. Walk through how you'd investigate this, and identify six concrete categories of defects that could cause false approvals, covering detection, fix, and prevention for each.

“This one took me a second to orient.”

View Post

Amazon·Software Engineer·Onsite - Behavioral / Leadership

Jun 2026
Amazon SWE behavioral round, basically one question about whether you've dropped the ball on something. Pretty standard leadership principles territory.
  • Can you describe a time when you failed to meet a commitment you had made?

“I had a real example ready but fumbled the delivery.”

View Post

Amazon·Software Engineer·Technical Phone Screen

Jun 2026
Amazon SWE coding round, basically a Top-K problem with some follow-up discussion on trade-offs. Pretty standard but the streaming variant caught me a bit flat-footed.
  • Given a large collection of items with associated scores or frequencies, return the K items with the highest values. Walk through your approach and discuss the trade-offs between different solutions.

“I started with the naive sort-and-slice answer, which they clearly expected, then moved to the min-heap approach.”

View Post

Amazon·Software Engineer·Online Assessment (OA)

Jun 2026
Amazon SWE interview with a system design coding problem around a parking lot. Pretty straightforward on the surface but there's enough edge cases to trip you up if you're not careful about how you model the space allocation logic.
  • Design a parking lot system that handles vehicle entry and exit for three car sizes (small, medium, large), allocates the correct space type per vehicle, and reports available spaces after each operation.

“My first instinct was to just use three counters and call it done, which is basically the right call here.”

View Post

Amazon·Software Engineer·Onsite - Behavioral / Leadership

Jun 2026
Behavioral loop at Amazon for a software engineer role, two questions that were basically the whole session. Both were ownership and delivery under pressure, which is pretty standard for them, but the depth they want is not.
  • Tell me about a time you took on responsibility that was clearly outside your job description. Walk through why you stepped in, how you got others aligned, how you handled the risks, and how you measured whether it actually worked.
  • Describe a situation where you had to deliver something important under a tight deadline with fewer resources than you needed. What was at stake, how did you decide what to cut, what was the result, and what did you take away from it?

“This one is sneaky because it sounds like a simple ownership story but they kept probing on the stakeholder alignment piece.”

View Post

Amazon·Software Engineer·Technical Phone Screen

Jun 2026
Security engineer screen at Amazon, just one question about SOC automation. Short and a bit underwhelming honestly.
  • Walk me through SOC processes you have automated and how you approached it.

“Decent question but the notes here are basically empty so I'm going off instinct.”

View Post

Amazon·Machine Learning Engineer·Technical Phone Screen

Jun 2026
Technical phone screen for an ML Engineer role at Amazon, focused entirely on LLM internals. Two topics: tokenization design and the SFT training objective. No behavioral questions, no coding, just "explain why it works this way" probing for about an hour.
  • Why do we tokenize text at all, and why do practitioners use subword schemes like BPE or unigram instead of whole-word or character-level tokenization?
  • Could you build a tokenizer using just the 26 letters of the English alphabet? Would it technically work, and what breaks in practice?
  • In supervised fine-tuning, can you add a KL divergence term to the loss? If so, what are the two distributions involved, what does the combined objective look like, and why would you want it?
  • Tokenizers often split numbers or code identifiers into multiple subword pieces in ways that hurt arithmetic and code reasoning. Why does this happen and what tokenizer-level choices can help?
  • The KL regularization term requires a second forward pass through a frozen reference model at every training step. How would you make this cheaper, and what do you give up by approximating it?
  • After SFT, the model follows instructions well but has clearly forgotten pretraining knowledge. How do you use the KL coefficient, data mixing, or reference checkpoint choice to diagnose and fix this?

“This felt like a warm-up but it actually has teeth.”

View Post

Amazon·Product Manager·Onsite - Product Sense / Strategy

Jun 2026
Amazon PM interview with a meaty product strategy case built around a Kindle launch scenario. The whole thing revolved around a classic scope vs. timeline tradeoff, and they made you work through it from multiple angles before asking you to actually make the call.
  • A new Kindle is four months behind schedule with eight planned features across software and hardware. What are the strongest arguments for delaying the launch to ship the full feature set?
  • Now flip it: what are the strongest arguments for shipping on time with a reduced feature set to hit the Black Friday window?
  • Given a feature breakdown table showing progress, user value, cost, and complexity for each of the eight features, decide which to cut and which to keep to make the Black Friday date. Walk through your prioritization logic in order of importance.

“I went straight to brand integrity and the risk of a half-baked product defining the device for reviewers.”

View Post

Amazon·Software Engineer·Onsite - System Design / Architecture

Jun 2026
System design round at Amazon for a software engineering role. The question was a beast, basically a full distributed systems design covering job scheduling at massive scale, and it went deep fast. I left feeling like I'd only scratched the surface on half the sub-topics.
  • Design a horizontally scalable, multi-tenant job scheduler service that supports ad-hoc and recurring jobs (cron and fixed-rate), with at-least-once execution, idempotency, and multi-region operation.
  • How would you efficiently find all jobs due in the next 5 minutes across millions of scheduled jobs?
  • Walk through the data model and schema for the scheduler, including how you'd handle partitioning and indexing for jobs, schedules, executions, and dead-letter records.
  • How would you handle multi-region operation and clock skew in a distributed job scheduler?
  • What observability and failure recovery mechanisms would you put in place for this system?
  • How do you handle per-tenant fairness and quota enforcement at scale, and what are the trade-offs between at-least-once and exactly-once delivery?

“This question ate me alive a little.”

View Post