← Back to Directory

Amazon interview notes, page 30 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·Software Engineer·Onsite - Behavioral / Leadership

Apr 2026
Amazon SWE behavioral loop, all three questions were leadership principle style. Nothing technically hard but I definitely underestimated how much prep these rounds need.
  • Tell me about a time you had a conflict with a teammate and how you resolved it.
  • Describe a situation where you took ownership of something outside your formal responsibilities.
  • Give an example of when you proactively learned something new to solve a problem at work.

“I had a story ready but it came out kind of flat.”

View Post

Amazon·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Amazon ML engineer screen, pretty focused on fundamentals. Just two questions but they wanted you to actually explain the reasoning, not just recite definitions.
  • What are the assumptions underlying linear regression?
  • Why do we use squared loss instead of other loss functions like absolute error?

“I listed the usual ones: linearity, independence of errors, homoscedasticity, normality of residuals.”

View Post

Amazon·Software Engineer·Recruiter / HR Screen

Apr 2026
Behavioral screen for a SWE role at Amazon. Just one question from what I can tell, and it was a pretty standard motivation question.
  • Why do you want to work at Amazon?

“Worth noting the content originally said Hinge, not Amazon, so either the notes got mixed up or someone copy-pasted the wrong thing.”

View Post

Amazon·Product Manager·Onsite - Behavioral / Leadership

Apr 2026
Amazon PM behavioral round, just one question but it took up most of the time. Pretty standard leadership principle territory but I fumbled the depth part more than I expected.
  • Describe a situation where you had to ask probing questions or investigate further to truly understand what was going on.

“I had a story ready but the interviewer kept pushing for more detail on why I chose to dig deeper versus just accepting the initial answer.”

View Post

Amazon·Technical Program Manager·Onsite - Multi Round

Apr 2026
Amazon TPM loop, mix of leadership principles, functional stuff, and one system design question at the end. Pretty standard format but the volume of behavioral questions back to back was a lot to keep track of.
  • Tell me about a time you demonstrated a strong sense of ownership over a project or outcome.
  • Describe a situation where you had to make a decision with incomplete information and push a team forward anyway.
  • Give an example of a time you raised the bar for quality on your team or held others to a higher standard.
  • How do you manage technical roadmap prioritization when engineering capacity is constrained and multiple stakeholders have competing priorities?
  • Walk me through how you track and communicate program health across multiple workstreams to senior leadership.
  • Design a system for tracking real-time inventory across a distributed network of fulfillment centers.

“Went with a story about a launch that almost slipped because a dependency team depooled.”

View Post

Amazon·Software Engineer·Technical Phone Screen

Apr 2026
SQL pivot question for a BI Engineer role at Amazon. Pretty focused technical screen, just the one problem but it had some depth to it if you weren't already comfortable with conditional aggregation.
  • You have a table with department, revenue, and month columns. Write a SQL query that transforms it so each department gets one row, with separate columns for each month's revenue (Jan_Revenue, Feb_Revenue, and so on through December).

“The PIVOT keyword slipped my mind mid-interview so I went with CASE statements inside SUM() which works fine but felt messier to write out loud.”

View Post

Amazon·Software Engineer·Onsite - Behavioral / Leadership

Apr 2026
Amazon software engineer behavioral round, one question deep enough that it basically ate the whole session. They wanted a real investigation story, not a vague 'I fixed a bug' answer.
  • Tell me about a time you had to independently investigate a complex problem that no one else had the capacity or knowledge to tackle, and drove it all the way to a concrete conclusion.

“This is the kind of question where a mediocre answer kills you.”

View Post

Amazon·Software Engineer·Technical Phone Screen

Apr 2026
Amazon SWE interview with a coding problem that's basically a twist on a classic LC problem. Not much else to go on but the question itself was interesting enough to remember.
  • Find the longest subsequence in an array where the difference between any two adjacent elements is less than k.

“It's the consecutive sequence problem but with a twist.”

View Post

Amazon·Software Engineer·Onsite - Behavioral / Leadership

Apr 2026
Amazon SWE interview with a scenario-based question about code review ethics. Pretty niche format where you rank actions from most to least effective rather than just describing what you'd do.
  • You're reviewing a colleague's code that has already been approved by another engineer, and you spot a bug that will affect customers. How do you rank the following options from most to least effective: ignore the issue, approve the code and patch it after deployment, consult the approving engineer before deploying, leave a comment on the review and withhold your approval, or document the issue for a post-launch fix?

“The ranking format threw me a bit.”

View Post

Amazon·Software Engineer·Online Assessment (OA)

Apr 2026
Amazon SWE coding round with a geometry-heavy simulation problem. The diagonal traversal angle was a bit unexpected and took some time to wrap my head around.
  • Given n servers on an infinite 2D plane and a sequence of directional redirects (NE, SE, NW, SW), simulate a traversal starting from the first server. Each redirect moves to the nearest unvisited server along the corresponding diagonal ray. If no such server exists, skip the step. Return the final server coordinates after processing all redirects.

“The diagonal constraint is what trips you up at first.”

View Post