Amazon Interview Questions

You have read the list of Leadership Principles and it probably sounded like wall art. This section explains what interviewers do with them, which is more concrete than the phrasing suggests. Each interviewer in your loop is assigned a subset of principles and has to leave the room with written evidence that you demonstrated them.

That assignment is why the same loop can feel repetitive. One interviewer is hunting for Customer Obsession and Ownership, another for Dive Deep and Bias for Action, so you get asked for a different story about broadly similar territory. Prepare a set of six to eight distinct stories rather than one favorite, and know which principles each story can serve.

• Customer Obsession. A decision you made that started from the customer, not from the roadmap.
• Ownership. Something outside your job description that you picked up because it was broken.
• Dive Deep. A problem where you went to the raw data and found what the summary hid.
• Have Backbone; Disagree and Commit. A time you pushed back, lost, and then executed anyway.

The phrasing varies by interviewer, but the shapes repeat across loops and levels.

• Tell me about a time you took on something significant outside your area of responsibility.
• Describe a time you disagreed with your manager or a peer. What did you do?
• Tell me about the most complex problem you have worked on. How did you find the root cause?
• Give an example of a time you had to make a decision with incomplete data.
• Tell me about a time you failed or missed a deadline. What changed afterward?
• Describe a time you simplified a process or cut cost without hurting quality.

Expect the follow-ups more than the questions. After your story, interviewers ask what your specific contribution was, what the numbers were before and after, what you would do differently, and what your teammates would say about it. A story that cannot survive four follow-ups reads as borrowed.

Use STAR, but weight it the way Amazon weights it: a short situation, one sentence of task, most of your time on action in the first person, and a result with a number in it. "We improved latency" fails. "I cut p99 latency from 800ms to 210ms by adding a read-through cache, which held through Prime traffic" passes.

Consider a supply chain analyst interviewing for a senior role. She brings a story about a forecasting fix and gets asked, in order, what her own model change was, why she chose that approach over two alternatives, what the error rate was before, and what broke first when volume tripled. She had the numbers, so the story held. Her second story did not have numbers, and the interviewer moved on within a minute.

Rehearsing this out loud is the part most candidates skip. You can practice the follow-up pressure against an AI interviewer on the /mock-interview page, which is where the difference between a written story and a spoken one shows up.

During a spoken video round, SubcueAI transcribes the interviewer's question and surfaces a suggested structure on your side, either in the desktop overlay on macOS and Windows or in the Chromium browser extension side panel. No meeting bot joins the call, and nothing is injected into the meeting page. For behavioral rounds the useful output is a reminder of which of your own stories fits the principle being probed, not a script; the numbers still have to be yours.

Two honest limits. If a technical round moves onto a screen-monitored or proctored coding platform, live help is out of scope and you should treat that round as your own skills only. If you are asked to share your screen, anything on your screen is visible. More on how major employers structure their loops sits in the company interviews topic, and role-specific question sets are collected under question banks.