Amazon India Interview Questions & Answers (2026)

Here's the thing nobody tells you about interviewing at Amazon: you can answer every technical question correctly and still get rejected.

That's not a scare story. It's how the process is designed. Amazon scores candidates against 16 Leadership Principles, and those principles aren't a poster in the lobby — they're the literal rubric your interviewer fills in after you leave the room. Each interviewer in your loop is assigned two or three specific principles to assess you on. If you're assigned Frugality and you have no story ready for it, a brilliant Customer Obsession answer will not save you.

Most candidates in India walk in having practised data structures for six weeks and the behavioural round for about an hour. That ratio is backwards.

This guide covers Amazon India interview questions and answers — the Leadership Principles, the behavioural questions that come from them, the technical rounds, and what the Bar Raiser is actually there to do. Each question tells you what the interviewer is checking, followed by an effective answer you can adapt to your own experience.

Amazon India Interview Process 2026

Amazon India hires across far more than software. The process shape changes by role, but the Leadership Principles run through all of them.

The main hiring locations are Bengaluru, Hyderabad, Chennai, Gurugram and Pune, with fulfilment-centre roles spread far wider.

The Rounds

1. Online assessment. For SDE roles, two coding problems plus work-style and work-simulation sections. For business and analytics roles, expect Excel, SQL or a case-based assessment. The work-style section is itself an LP test — answer as the Leadership Principles would, not as you'd answer a personality quiz.
2. Phone screen or first technical round. Usually one interviewer, one or two problems, plus LP questions layered on top.
3. The loop. Three to four back-to-back interviews. Every one of them contains behavioural questions. There is no round where the Leadership Principles stop mattering.
4. The Bar Raiser. One interviewer in your loop is from a completely different team, trained specifically to keep hiring standards consistent. Their job is to ensure every hire raises the average — and they hold veto power. A Bar Raiser can block an offer even when the hiring manager wants you.

You won't be told which interviewer is the Bar Raiser. Assume every round is being scored equally hard, because it is.

The 16 Amazon Leadership Principles

Amazon has 16 Leadership Principles. The list moved from 14 to 16 in 2021 and hasn't changed since — if a prep guide you're reading lists 14, it's out of date.

The mistake almost everyone makes: over-preparing Customer Obsession and Deliver Results while ignoring Frugality, Learn and Be Curious and Have Backbone. Those are the ones candidates get assigned and freeze on.

The fix, build a story bank. Prepare five to six detailed stories from your work, projects or internships, each with the numbers attached. A single strong story usually covers three or four principles depending on which angle you tell it from. That's far more effective than trying to memorise 16 separate answers.

How to Structure Every Behavioural Answer: STAR

Amazon interviewers are trained to score STAR-format answers, and they will interrupt to extract the missing pieces if you skip one.

• Situation — the context, in two sentences maximum
• Task — what you specifically were responsible for
• Action — what you did, not what the team did
• Result — what changed, with a number

Two failures cost candidates more than anything else. The first is saying "we" throughout the interviewer cannot score a team. The second is stopping before the result. Amazon is a results culture; an answer without an outcome reads as incomplete. If you don't have exact figures, use a range or a direction: "processing time dropped from about two days to a few hours."

Amazon Leadership Principles Interview Questions & Answers

1. Tell Me About a Time You Went Above and Beyond for a Customer

Customer Obsession. The interviewer wants evidence you'll take on inconvenience for a customer who isn't in the room. The weak version of this answer describes doing your job well. The strong version involves a cost to you.

Effective Answer: "During my internship I handled support tickets for a small logistics product. A client reported that their weekly report was arriving with duplicated rows. The quick fix was to tell them to filter it in Excel, which is what the previous ticket had done. I checked the data and found the duplication came from a sync running twice on days when a job retried. I flagged it, wrote the deduplication logic, and tested it against three months of their historical files. It took two days I hadn't planned for. The client stopped raising that ticket entirely, and two other accounts on the same job had been silently affected too."

2. Describe a Time You Took Ownership of Something Outside Your Responsibility

Ownership. Amazon is explicitly looking for people who don't say "that's not my job." Pick something you weren't asked to do.

Effective Answer: "Our final-year project had four people and no version control — everyone was mailing files around, and we lost two days of work when someone overwrote a module. Nobody owned this and it wasn't my assigned area. I set up a GitHub repo over a weekend, wrote a one-page guide for the two teammates who hadn't used Git, and moved everything across. We didn't lose work again, and by the review we had a clean commit history that our guide specifically commented on."

3. Tell Me About a Time You Simplified a Complex Process

Invent and Simplify. Note that simplification scores as highly as invention here. Removing steps counts.

Effective Answer: "Our college fest registration ran on a Google Form, and three volunteers spent about six hours a day copying entries into a spreadsheet and mailing confirmations. I replaced it with a form that wrote directly into a sheet, with a script that sent the confirmation automatically. It wasn't clever engineering — it removed a manual step that existed only because nobody had questioned it. The six hours a day went to roughly twenty minutes of checking, and we handled about 1,200 registrations without a single missed confirmation."

4. Tell Me About a Time You Made a Decision With Incomplete Information

Bias for Action, and often Are Right, A Lot. Amazon values calculated speed. Show that you knew what you didn't know and moved anyway.

Effective Answer: "Two days before our project demo, our hosting provider started timing out intermittently. I couldn't diagnose whether it was them or our code in the time available. Rather than spend the last two days debugging something I might not control, I deployed a duplicate instance on a second provider and pointed the demo at whichever was responding. It cost me an evening and about ₹400. The original provider was fine on the day, so the backup went unused — but I'd have made the same call again, because the downside was one evening and the upside was not failing the demo."

5. Tell Me About a Time You Disagreed With Your Manager or a Teammate

Have Backbone; Disagree and Commit. This is a two-part question and most candidates answer only the first half. You must show the disagreement and the commitment afterwards — including when you were overruled.

Effective Answer: "During my internship I disagreed with my mentor about adding a caching layer for a feature. I thought the data changed too often for the cache to help and that it would cause stale-data bugs. I put together a short note with the update frequency from the logs and shared it rather than arguing in the standup. He still went ahead — his reasoning was that a related feature would need the same cache within a quarter, which I hadn't known. I implemented it properly rather than half-heartedly, and I added an invalidation hook for the case I'd been worried about. It turned out he was right about the roadmap."

6. Tell Me About the Biggest Mistake You've Made

Earn Trust. The trap is a fake mistake. Interviewers are trained to spot candidates who nominate something harmless. Pick something that actually cost something, and be specific about what you changed afterwards.

Effective Answer: "I ran a delete query on our project database without a WHERE clause during testing and wiped about 400 records. We had a backup from four days earlier, so we lost four days of test data and I lost a full day recreating it. Two things came out of it. I never again run a delete without running it as a SELECT first to see exactly what it will touch. And I set up an automated daily backup, which we hadn't had, because the real failure wasn't my typo — it was that a single typo could cost four days."

7. Tell Me About a Time You Had to Learn Something Quickly

Learn and Be Curious. One of the most commonly assigned and least prepared principles. Show self-directed learning, not a course you were made to take.

Effective Answer: "Our project needed a dashboard and nobody on the team knew any visualisation tool. I gave myself a week on Power BI — official documentation for the data model, then rebuilt a sample report three times until I understood why the relationships mattered rather than just clicking through. The dashboard shipped in the final build. More usefully, understanding the data-model side meant I could tell when a chart was misleading, which happened twice with our own data."

8. Tell Me About a Time You Delivered a Result Under a Tight Deadline

Deliver Results. Amazon wants to see prioritisation, not heroics. An answer built on working all night scores worse than one built on cutting scope intelligently.

Effective Answer: "Our review was moved up by ten days with two modules unfinished. I listed everything outstanding and split it by what the reviewers would definitely test versus what was optional. We cut two features nobody had asked for and finished the escalation logic properly instead of half-finishing four things. We presented four days early with everything working. The lesson I took was that under time pressure the first decision is what to stop doing."

9. Give Me an Example of Doing More With Less

Frugality. Rarely prepared, frequently assigned. Constraints can be budget, headcount, time or tools.

Effective Answer: "We needed SMS notifications for our project and the paid gateways were outside a student budget. Instead of dropping the feature, I checked what our users actually needed — they wanted to know their complaint status had changed, not the specific detail. I built it on email plus a browser notification, both free, and kept SMS as a documented future upgrade. We delivered the outcome for zero cost rather than dropping it because the obvious solution was unaffordable."

10. Tell Me About a Time You Dug Into Data to Find a Problem

Dive Deep. Expect follow-ups pushing for exact numbers. This is where candidates who exaggerated earlier get caught.

Effective Answer: "Our registration numbers looked healthy but attendance was around 40% of registrations. Everyone assumed students were simply not showing up. I broke the data down by registration date and found that most no-shows had registered in the first three days, when our form had a bug that didn't send a confirmation email. The people who never got a confirmation forgot. We resent confirmations to that group and attendance in that cohort went up substantially."

11. Tell Me About a Time You Improved Something Nobody Asked You to Improve

Insist on the Highest Standards, often with Ownership.

Effective Answer: "Our API had no error messages — every failure returned a generic 500, and debugging meant reading logs line by line. It worked, so it wasn't on anyone's list. I spent a day adding specific error responses and status codes. It saved me time immediately, and when a teammate hit a validation failure two weeks later he fixed it himself in ten minutes instead of asking me. The standard I'd argue for is that working isn't the same as finished."

12. Where Do You See Yourself in Five Years?

Standard, but at Amazon it's read through Think Big and Learn and Be Curious. Avoid a title-ladder answer. Talk about scope of problem instead.

Effective Answer: "I'm less attached to a title than to the size of problem I'm trusted with. In five years I'd like to be owning a service or a workflow end to end — responsible for whether it works, not just for the piece I was handed. Realistically that means a couple of years getting genuinely strong at the fundamentals first. What draws me to Amazon specifically is that ownership seems to be given early rather than earned through tenure."

13. Why Amazon?

whether you understand the culture you're applying into, or just the brand and the pay. Reference the operating model, not the logo.

Effective Answer: "Two things. The scale of the problems Amazon India is solving delivery and logistics problems in a market where addresses aren't standardised and infrastructure varies enormously by city. Those are genuinely harder problems than the same company faces in the US, and they're being solved here. Second, the ownership culture. I've read enough about how Amazon works to know that people are expected to own outcomes rather than tasks, and be held to it. That suits how I want to work, and honestly it's the part I'd find hardest, which is also why I want it."

Amazon India Technical Interview Questions

Technical depth varies enormously by role. SDE candidates face serious data structures and algorithms; analytics roles face SQL and case work; operations roles face scenario and metric questions. What stays constant is that Leadership Principles questions are woven into these rounds too.

14. What is the Difference Between an Array and a Hash Map, and When Would You Use Each?

Tests whether you choose data structures deliberately. Give the complexity and the decision rule.

Effective Answer: "An array gives O(1) access by index and stores elements contiguously, which makes it cache-friendly for iteration. A hash map gives roughly O(1) lookup by key rather than position, at the cost of extra memory and no inherent ordering. If I know the position I need, array. If I'm looking things up by an identifier — a user ID, an order number — hash map. Most of the optimisation problems I've solved came down to replacing a nested loop with a hash map lookup."

15. How Would You Find the First Non-Repeating Character in a String?

A very common Amazon screening problem. Talk through the approach before writing, and state the complexity unprompted.

Effective Answer: "Two passes with a hash map. First pass counts the frequency of each character. Second pass walks the string in order and returns the first character with a count of one. That's O(n) time and O(k) space where k is the character set size. The naive approach for each character, scan the rest of the string is O(n²), which matters as soon as the input gets large. Edge cases: empty string, and a string where every character repeats, both returning a defined sentinel."

16. Explain SQL JOINs and When You'd Use Each

Essential for analytics, business analyst and brand specialist roles in India.

Effective Answer: "A JOIN combines rows across tables on a related column. INNER JOIN returns only matching rows. LEFT JOIN returns everything from the left table with nulls where the right has no match which is what you use when you need to find what's missing, like customers with no orders. RIGHT JOIN is the mirror. FULL OUTER JOIN returns both sides. In practice, most business questions are answered with INNER and LEFT."

17. How Would You Investigate a Sudden Drop in a Key Metric?

Asked in analytics and operations interviews, and it's really a Dive Deep test in technical clothing. Structure beats intuition here.

Effective Answer: "First I'd confirm the drop is real rather than a reporting issue — check whether the pipeline ran, whether the definition changed, whether it's one source or all of them. Then segment: by region, by device, by customer type, by time of day. A metric that drops uniformly points at something systemic; a metric that drops in one segment points at something local. Then look for what changed at the same time — a release, a pricing change, a competitor event, a festival calendar shift. Only then do I form a hypothesis, because the segmentation usually names it for you."

18. How Would You Handle an Underperforming Associate? (Operations / Area Manager roles)

Amazon's fulfilment-centre roles in India are people-management jobs. This tests Hire and Develop the Best alongside Earn Trust.

Effective Answer: "First I'd find out whether it's a can't or a won't, because the response is completely different. That means looking at their actual numbers against the team's rather than going on impression, then having a direct conversation. If it's a skill or process gap, that's on me to fix retraining, pairing them with a strong performer, checking whether their station or shift is the real problem. If it's engagement, I'd want to understand what changed. I'd set a specific expectation with a timeline and document it, so the person knows exactly what good looks like rather than just feeling watched."

5 Mistakes That Get Candidates Rejected at Amazon

• Preparing technically and winging the behavioural round. This is the single biggest cause of rejection for strong engineers. The LPs are the rubric.
• Saying "we" instead of "I". Interviewers cannot score a team's achievement. They need your specific action.
• No numbers. An answer that ends without a result reads as unfinished in a results culture.
• Only preparing the famous principles. If you're assigned Frugality or Have Backbone and have nothing ready, that's a scored gap.
• Inflating a story. Dive Deep follow-ups will ask for the specifics. A story you exaggerated collapses under three questions, and losing trust is worse than a weak example.

How to Prepare: A 3-Week Plan

Week 1 — Build the story bank. Six detailed stories from projects, internships or work, written out in STAR with real numbers. Then map each against all 16 principles and find your gaps. Most people discover they have nothing for Frugality and nothing for Have Backbone.

Week 2 — Role-specific technical. SDE: arrays, strings, hash maps, trees, and complexity analysis out loud. Analytics: SQL joins, window functions, and case structure — our free SQL Mastery crash course covers the query patterns that recur. Operations: metrics, scenarios and people situations.

Week 3 — Rehearse out loud, with follow-ups. Have someone ask you "what was the number?", "why did you choose that?", "what would you do differently?" after every story. That's the Bar Raiser experience in miniature, and it's where under-prepared stories fall apart.

Final Thoughts

If you take one thing from this guide, take this: Amazon is not primarily testing whether you can code. It's testing whether you behave the way Amazon has decided good employees behave, and it has written down exactly what that looks like in 16 lines.

That's unusually generous, when you think about it. Most companies make you guess what they want. Amazon publishes the rubric.

So build the story bank. Put real numbers in it. Say the stories out loud until the follow-up questions stop surprising you. Then go in knowing what's being scored.