← Back to Directory

Amazon interview notes, page 73 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·Machine Learning Engineer·Technical Phone Screen

May 2026
Amazon MLE loop, three questions total. Two coding problems and one ML concept explanation. Nothing too wild but the K-means one took more time than I expected to articulate cleanly.
  • Explain or implement the K-means clustering algorithm, covering initialization, assignment, centroid update, and stopping criteria. Make sure dimensions are consistent throughout.
  • Given a list of closed intervals, merge all overlapping ones and return the result.
  • Given an array of integers and a value k, return the k most frequently occurring elements.

“I know K-means cold, or so I thought.”

View Post

Amazon·Software Engineer·Onsite - Coding / Algorithms

May 2026
Amazon SWE coding round focused on a classic array problem but pushed well beyond the surface. They wanted the full O(n) no-division solution plus a live dry run and edge case reasoning, which made it more involved than I expected.
  • Given an integer array, return a new array where each element is the product of all other elements in the original array. No division allowed, must run in O(n) time and use O(1) extra space beyond the output array.
  • How does your solution behave with arrays containing zeros or negative numbers?
  • How would you handle integer overflow in a language with fixed-width integers when computing these products?

“I knew the prefix/postfix trick going in, which helped, but the no-division constraint still trips people up if they haven't seen it before.”

View Post

Amazon·Software Engineer·Technical Phone Screen

May 2026
Amazon SWE interview with a scenario-based question about balancing speed vs. quality under stakeholder pressure. Just one scenario but it had a lot of moving parts and I don't think I ranked the strategies as cleanly as they wanted.
  • A customer-facing redirect button is broken in production. Your PM wants a fix by end of day, your engineering manager wants 4 days to do it properly with unit tests. You're given several testing strategies and asked to rank them by effectiveness.

“The strategies ranged from things like rolling back the change and retesting from scratch, to spot-checking a subset and just shipping it, to writing new targeted unit tests.”

View Post

Amazon·Software Engineer·Onsite - Behavioral / Leadership

May 2026
Behavioral round at Amazon for a SWE role. Just the one question but it hit hard, classic Amazon pressure-test stuff.
  • Describe a situation where you had to make a decision quickly despite having many possible options.

“I fumbled the opening on this one.”

View Post

Amazon·Software Engineer·Technical Phone Screen

May 2026
Amazon data engineering interview with a system design question around large-scale CSV deduplication. Pretty technical, felt like a classic scalability probe dressed up as a data problem.
  • You have a large set of CSV files, each containing thousands of paragraphs. How would you detect duplicate paragraphs within a single file, and then scale that approach across many files?

“I jumped straight to hashing, which felt right, but I didn't think through the multi-file case fast enough.”

View Post

Amazon·Machine Learning Engineer·Technical Phone Screen

May 2026
Amazon ML engineer interview, came down to a technical phone screen focused on model fundamentals. Pretty standard stuff but the loss function question had more depth to it than I expected.
  • What are some effective loss functions used in computer vision models?

“I started with cross-entropy for classification and MSE for regression tasks, then moved into stuff like focal loss for class imbalance and IoU-based losses for object detection.”

View Post

Amazon·Software Engineer·Onsite - Coding / Algorithms

May 2026
Amazon SWE live coding round where they had me build a web crawler from scratch in my own IDE with screen sharing. Open-book but no AI tools allowed, which honestly felt fair. The problem sounded straightforward until I started actually writing it.
  • Implement a BFS-based web crawler starting from a seed URL, with URL normalization, deduplication, and an option to restrict crawling to the same domain.
  • How would you handle HTTP errors, malformed URLs, and timeouts in your crawler, and what retry logic would you implement for transient failures?

“I started with the BFS queue and dedup set, which went fine.”

View Post

Amazon·Software Engineer·Technical Phone Screen

May 2026
Amazon Applied Scientist phone screen, one coding question the whole time. Pretty standard probability/sampling problem but the follow-up on complexity tripped me up a bit.
  • Design a class that takes an array of weights and supports a method that returns a random index where each index is picked with probability proportional to its weight.

“I knew prefix sums were the move but fumbled explaining why you binary search instead of just scanning linearly.”

View Post

Amazon·Software Engineer·Technical Phone Screen

May 2026
Amazon SWE interview with an OOD round. Just the one design question but it took up the whole session, lots of back and forth on the details.
  • Design a publish-subscribe (pub-sub) system from an object-oriented perspective.

“Jumped straight into defining a Publisher and Subscriber interface, which felt right, but I fumbled when they pushed on how you'd handle multiple topics and fan-out.”

View Post

Amazon·Software Engineer·Online Assessment (OA)

May 2026
Took the Amazon SDE online assessment and came out with mixed feelings. The behavioral and work style sections were fine, the AI coding question went clean, but the LeetCode problem hit a wall with 5 test cases failing to TLE. Knew the fix, just couldn't code it in time.
  • Algorithmic coding problem with performance requirements, where a brute-force solution passes some but not all test cases due to time limit constraints.

“Passed 10 of 15.”

View Post