You are interviewing for a **Software Engineer (Intern)** role at **Amazon**, in an on-site loop of two back-to-back 60-minute rounds. Each round mixes a behavioral block with a coding problem; the prompts below are the behavioral portion. One round was with a peer engineer, the other with the hiring manager. Answer each prompt using a clear, concrete example from your past work, projects, internships, research, or coursework — one story per prompt, with your personal contribution front and center. Amazon explicitly scores behavioral answers against its **Leadership Principles** (e.g., *Ownership*, *Bias for Action*, *Earn Trust*, *Dive Deep*, *Deliver Results*, *Learn and Be Curious*, *Are Right, A Lot*), so each story should surface evidence for the principle its prompt is testing. --- ### Constraints & Assumptions - This is an **early-career / intern** loop: interviewers expect honest examples from school, internships, side projects, research, or hackathons — not necessarily large production systems. - **Format:** two back-to-back 60-minute rounds; budget each behavioral answer to roughly **2–4 minutes** spoken, leaving room for follow-up drilling and the coding question in the same round. - Amazon expects **specifics and metrics**. Vague stories ("we worked hard and shipped it") fail; concrete tools, constraints, tradeoffs, and numbers pass. - Assume the interviewer will interrupt with "what was *your* part?" and "what did the data show?" — your story must survive that probing, so it should be true and your own. ### Clarifying Questions to Ask For a "tell me about a time" prompt there is usually little to clarify — start telling the story. The one or two worth a quick check up front, before you begin: - Would you prefer a story from a professional/internship setting, or is academic, research, or personal-project work equally welcome? - Are you looking for the most impactful example I have, or one that best fits the specific principle this round is probing? --- ### Part 1 — A tight deadline Tell me about a time you faced a tight deadline. What was at stake, how did you decide what to do, and what was the outcome? ```hint Structure Use a structured narrative (Situation → Task → Action → Result). The signal lives in the **Action**: what you cut, parallelized, or escalated — not that you "worked hard." ``` ```hint What good looks like Show a deliberate **tradeoff under constraint** — scope reduction, prioritization, surfacing risk early — rather than heroics. Tie it to *Bias for Action* and *Deliver Results*. ``` #### What This Part Should Cover - **A genuinely hard constraint** — a fixed external date and scope larger than the time allowed, not just self-imposed busyness. - **Deliberate prioritization** — what was cut, deferred, or parallelized, and the reasoning behind it. - **Early risk communication** — surfacing the squeeze to the right people rather than absorbing it silently. - **A quantified result** — what shipped on time and the measurable value it delivered. ### Part 2 — A mistake you made Tell me about a time you made a mistake. How did you discover it, what did you do about it, and what changed afterward? ```hint Pick the right story Choose a **real** mistake with genuine consequence that you **owned** — not a disguised humble-brag ("I care too much"), and not someone else's fault. Accountability is the point. ``` ```hint Land the ending The recovery and a **systemic prevention** (a test, a check, monitoring, a process change) matter more than the slip itself. This maps to *Earn Trust* and *Ownership*. ``` #### What This Part Should Cover - **Plain accountability** — the mistake stated without hedging or blame-shifting. - **Fast detection and mitigation** — how you noticed it and limited the damage. - **Transparent communication** — proactively telling the affected people rather than hiding it. - **A systemic fix** — a durable prevention (test, guardrail, monitoring, process) rather than "I'll be more careful." ### Part 3 — A difficult problem you solved Tell me about a time you solved a difficult problem. Explain how you **discovered** the problem, how you **developed** a solution, and how you **drove it to completion**. ```hint Cover all three verbs The prompt explicitly asks for discovery → solution → completion. Map your story to all three: how you *noticed* it (a metric, a bug report, a failing test), how you *chose* among alternatives, and how you *shipped and verified* the fix. ``` ```hint Show depth This is where *Dive Deep* and *Are Right, A Lot* are scored. Name the hypotheses you tested and the evidence that confirmed the root cause — not just the final fix. ``` #### What This Part Should Cover - **A concrete trigger for discovery** — a metric regression, bug report, failing test, or anomaly you chose to investigate. - **Evidence-driven root-cause analysis** — the hypotheses tested and the data that confirmed the actual cause. - **A justified choice among alternatives** — why your approach beat the options you considered. - **Verification after the fix** — how you confirmed it worked, plus a measurable improvement. ### Part 4 — Using generative AI tools Tell me about your experience using generative AI tools. Describe how you used them **responsibly** and what **impact** they had. ```hint Frame the boundary Position AI as an **assistant under your judgment**, not a replacement for it. Concretely: how did you *verify* the output (tests, review, reasoning) before trusting it? ``` ```hint Responsibility signals Name the real risks you managed — hallucinations, secret/confidential-data leakage, security, licensing — and how your usage respected them. This maps to *Learn and Be Curious* plus good judgment. ``` #### What This Part Should Cover - **A concrete use case** — a specific task (test generation, scaffolding, explaining an error, design brainstorming), not "I use it sometimes." - **You stayed accountable** — the AI assisted; you owned correctness and security. - **Explicit verification** — how you checked the output (ran it, reviewed/wrote tests, cross-checked docs) before trusting it. - **Risk boundaries respected** — privacy/confidentiality, security, licensing, and hallucination awareness, plus an honest impact. --- ### What a Strong Answer Covers These dimensions span all four parts; the per-Part rubrics above cover what each individual story must surface. - **Clear personal ownership** — "I" does the work in the story, even within a team effort; your specific contribution is unambiguous. - **Tight STAR structure** — a coherent narrative the interviewer can follow without re-asking, with most time spent on Action and Result rather than setup. - **Specificity and evidence** — concrete numbers, named tools/techniques, and the data that informed each decision. - **Outcome plus reflection** — a measurable (or at least observable) result, and an honest statement of what you learned or would do differently. - **Leadership-Principle alignment** — each story naturally demonstrates the relevant principle (deadline → *Deliver Results*; mistake → *Earn Trust* / *Ownership*; hard problem → *Dive Deep*; AI → judgment + *Learn and Be Curious*) without name-dropping it. ### Follow-up Questions - "What would you do differently if you faced that same situation today?" - "What was the hardest tradeoff in that decision, and who disagreed with you?" - "How did you know your fix actually worked — what did you measure?" - "Where did the generative-AI tool get something wrong, and how did you catch it?"