Amazon Software Engineer Interview Questions (2026)

How many rounds are in the Amazon software engineer interview loop?

Most candidates go through four stages: an online assessment (OA), a recruiter or hiring-manager phone screen, and a virtual onsite loop of four to five rounds. Each onsite round runs 60 minutes and contains both a behavioral segment (15–25 minutes, two Leadership Principle questions) and a technical segment (coding or system design). The full process takes four to eight weeks from OA to offer.

What is Amazon's online assessment format for SDEs?

The SDE OA is a single proctored session of 2–3.5 hours. It includes two algorithm coding problems (70–90 minutes combined), a Work Simulation module (~50 minutes presenting realistic workplace scenarios mapped to the Leadership Principles), and a Work Style Assessment (15–20 minutes). Amazon expects syntactically correct, original code — the platform flags copy-paste behavior.

What are Amazon's 16 Leadership Principles and why do they matter?

Amazon's 16 LPs are: Customer Obsession, Ownership, Invent and Simplify, Are Right A Lot, Learn and Be Curious, Hire and Develop the Best, Insist on the Highest Standards, Think Big, Bias for Action, Frugality, Earn Trust, Dive Deep, Have Backbone Disagree and Commit, Deliver Results, Strive to be Earth's Best Employer, and Success and Scale Bring Broad Responsibility. Every interviewer in your loop is assigned 2–3 specific LPs to probe. A full four-round loop generates 8–10 behavioral questions, and a weak LP signal can override a strong technical score.

What is the Bar Raiser and how much power do they have?

The Bar Raiser is a senior Amazonian from a completely different organization who is trained to maintain hiring quality. They hold veto power over the entire hiring committee — even if every other interviewer votes 'hire,' the Bar Raiser can block the offer. Their round is almost entirely behavioral, focused on LP depth and long-term fit, not raw coding ability. You will not know in advance which round is the Bar Raiser.

What coding topics does Amazon focus on?

Amazon coding rounds center on arrays and strings, hash maps, trees and graphs (DFS/BFS), dynamic programming, heaps and priority queues, and recursion. Questions are medium-to-hard LeetCode difficulty. Two problems per 70-minute coding round is standard; you can preview both and switch between them freely. Space/time complexity analysis is required — stopping at a brute-force solution without optimizing is a common 'no hire' signal.

Does Amazon ask system design questions for all engineer levels?

No. SDE I (L4, new grad or 0–3 years experience) loops typically include no system design or one light object-oriented design question. SDE II (L5) loops include one full system design round. SDE III (L6) and above get one or two design rounds and are expected to address scale, data partitioning, caching strategies, and trade-offs between consistency and availability.

What total compensation can I expect as an Amazon software engineer?

Based on Levels.fyi data, median total compensation is approximately $189K–$220K at L4 (SDE I), $271K at L5 (SDE II), and $390K at L6 (SDE III). Amazon's L5 TC trails Google L5 ($423K median) and Meta E5 ($480K median), but Amazon offers faster leveling for strong performers, and AWS teams often have above-median equity packages.

How should I structure a Leadership Principles answer at Amazon?

Use STAR format (Situation, Task, Action, Result), but Amazon interviewers probe deeper than other companies. After your answer they will ask 'What would you do differently?' and 'Tell me more about your specific contribution' to separate individual impact from team impact. Lead with the result in your opening sentence so interviewers can calibrate whether the scope is worth continuing, then walk backwards through the story. Quantify everything: lines of code reduced, latency improved, revenue protected, deadline met.

How long does the Amazon SDE hiring process take end to end?

Four to eight weeks is typical. The OA invite arrives within one to two weeks of applying. Phone screen scheduling takes one to two weeks after passing the OA. The onsite loop is usually condensed into one day (virtual). Hiring committee debrief and offer generation take two to five business days after the loop. Offer acceptance deadlines are typically five to seven business days, though extensions are usually granted if you ask.

How do I prepare for Amazon's behavioral rounds if I have limited industry experience?

Build a bank of five to six stories from academic projects, internships, or side work. Each story should demonstrate a measurable outcome and map to at least two LPs. Prioritize stories that show disagreement-then-commitment, data-driven decisions, or working with constraints — these cover the most-tested LPs (Ownership, Bias for Action, Have Backbone, Dive Deep, Deliver Results). Practice out loud with a timer; Amazon interviewers interrupt stories that run past four minutes.

Getting a software engineering role at Amazon is unlike any other FAANG process. The technical bar is high, but what uniquely defines Amazon’s loop is its Leadership Principles framework — a structured behavioral evaluation that runs through every round and carries as much hiring weight as your algorithm performance. This guide covers the 2026 loop from start to offer, what Amazon uniquely evaluates, real question types by round with sample answers, level and compensation context, and a concrete six-week prep plan.

The Amazon interview loop from application to offer

The process has four distinct stages. Treating any of them as a formality is a common way candidates lose offers they technically earned.

1. Recruiter screen (20–30 minutes)

The recruiter confirms basic fit: current role, years of experience, timeline, and target level. Be explicit about the level you want. Amazon levels SDEs into L4 (SDE I), L5 (SDE II), L6 (SDE III), and above. If you have five years of experience and allow the recruiter to default you to an L4 slate, you may be assessed against a lower bar and offered less — the leveling conversation is easier before the loop starts than after.

The recruiter will also send you a brief document about the Leadership Principles if you have not already encountered them. Read it. This is not optional background reading; it is the scoring rubric for 50% of every round you are about to take.

2. Online assessment (2–3.5 hours)

Amazon’s OA is a proctored, browser-based session with three or four components:

• Coding section: Two algorithm problems with a combined 70–90 minute window. You can preview both problems before starting and switch between them freely. Languages supported include Python, Java, C++, C#, JavaScript, and others. The platform monitors for copy-paste patterns and expects code written in real time.
• Work Simulation: ~50 minutes of scenario-based questions presenting realistic workplace situations. Each scenario is mapped to specific Leadership Principles. There is no objectively right answer, but there are patterns that signal LP alignment versus LP confusion.
• Work Style Assessment: 15–20 minutes of agree/disagree statements about work preferences. Inconsistent responses (saying you prefer fast decisions in one section and exhaustive analysis in another) will flag.

Passing the OA moves you to the phone screen. Roughly 30–40% of OA takers advance.

3. Phone screen (45–60 minutes)

Usually conducted by a hiring manager or senior engineer. The format mirrors a single onsite round in miniature: 15 minutes of behavioral questions covering one or two LPs, then 30–40 minutes on a coding problem. The coding environment is a shared browser editor (often CoderPad or an internal tool). You will be asked to analyze time and space complexity after your solution and to optimize it if your initial approach is suboptimal.

Treat this exactly like an onsite round. Candidates who coast through the phone screen and “save it” for the loop often discover that poor phone screen scores influence how aggressive the hiring committee is at debrief.

4. Virtual onsite loop (four to five rounds, one day)

This is the core of the process. All rounds run on the same day, back to back, via video call. Each round is 60 minutes. Interviewers score independently — they do not see each other’s scores until the post-loop debrief.

A typical L5 onsite loop looks like this:

L4 loops typically drop the system design round or replace it with an object-oriented design question. L6 loops add a second design round and expect you to drive the scope definition yourself.

What Amazon uniquely evaluates: Leadership Principles in depth

Every other FAANG company has behavioral interviews. Amazon is the only major tech company where behavioral questions are assigned, tracked by LP, and scored on a structured rubric across all five interviewers in a loop. Here is what makes it different in practice.

LP assignment per interviewer: Before the loop, each interviewer is assigned two to three specific LPs to evaluate. This means the panel collectively probes most or all 16 LPs without asking duplicate questions. Customer Obsession, Ownership, Deliver Results, and Bias for Action appear most frequently. Are Right A Lot, Frugality, and Hire and Develop the Best appear less often but are not rare.

The follow-up is the test: Amazon LP interviews are not pass/fail on your opening answer. They are scored on how deep you can go. After your story, every interviewer will follow up with at least two of these probes:

• “What was your specific contribution versus your team’s?”
• “What would you do differently today?”
• “How did you measure success?”
• “What data did you use to make that decision?”
• “What was the impact on the customer?”

A strong opening story that collapses under follow-ups scores lower than a modest story with rich, consistent depth.

Hypothetical answers fail: Amazon interviewers are trained to reject hypotheticals. If you say “I would…” instead of “I did…”, you will be redirected. If you run out of real examples, you will be scored lower even if your hypothetical reasoning is excellent. Build enough real stories before your loop that you never need to speculate.

The Bar Raiser: what it is and how to handle it

The Bar Raiser is an Amazon employee from a completely different organization — if you are interviewing for Prime Video, the Bar Raiser might be from AWS or Alexa. They are trained and certified specifically for this role, and they hold veto power over the hiring committee. A unanimous “hire” vote from four interviewers cannot override a Bar Raiser “no hire.”

Their job is not to make the loop harder. It is to ensure that every hire raises Amazon’s talent bar — meaning the candidate must be demonstrably stronger than at least 50% of people currently in that role at Amazon. You will not be told which round is the Bar Raiser round. Look for the interviewer who pushes hardest on depth, challenges your assumptions, and asks what you would do if your approach failed.

The Bar Raiser round is mostly behavioral. A typical structure:

• Two LP questions with aggressive follow-up (20–25 minutes)
• A coding problem or technical discussion (25–30 minutes)
• Questions about your career trajectory and long-term goals (5–10 minutes)

The LP depth here is more important than the coding. Bar Raisers have seen thousands of loops and can spot a rehearsed surface-level answer quickly. Give them real stories with the texture of what actually happened — the parts where things went wrong, the trade-offs you made, and the decisions you pushed back on.

Coding rounds: real question types and a sample approach

Amazon coding questions are medium-to-hard LeetCode difficulty, skewing toward data structures and algorithmic thinking over pure math. Topics that appear most frequently:

• Arrays and strings: sliding window, two-pointer techniques, in-place manipulation
• Hash maps: frequency counting, group anagram variants, two-sum family problems
• Trees and graphs: level-order traversal, DFS/BFS, detecting cycles, lowest common ancestor
• Dynamic programming: 0/1 knapsack, longest common subsequence variants, staircase problems
• Heaps: k-closest, median of stream, merge k sorted lists
• Linked lists: reversal, cycle detection, merge problems

Sample question (medium difficulty): Given an array of integers nums and a target integer k, return the number of contiguous subarrays whose product is less than k.

Approach narration (what Amazon interviewers want to hear): Start by confirming edge cases out loud: what if k is 1 or less? What if the array is empty? Then describe the brute-force O(n²) approach before jumping to the optimal solution — this shows you understand the space of solutions. The optimal approach is a sliding window with two pointers: maintain a running product, expand the right pointer, shrink the left when the product exceeds k. Count right - left + 1 subarrays at each step. Time complexity O(n), space O(1). Explain this before coding, code it cleanly, trace through an example, then state the complexity again at the end.

Amazon interviewers specifically look for: talking while coding (not silent), clean variable naming, edge case awareness before coding rather than after, and a willingness to propose a brute force before optimizing.

System design rounds (L5 and above)

System design rounds at Amazon run 45–60 minutes. For L5, one interviewer typically leads. Common question stems:

• “Design a URL shortener at scale”
• “Design Amazon’s product recommendation system”
• “Design a distributed notification service”
• “Design a real-time order tracking system”

What L5 interviewers evaluate:

1. Requirement clarification (2–4 minutes): Do you ask about scale, read/write ratio, latency requirements, consistency needs?
2. High-level design (10 minutes): Can you draw a coherent end-to-end architecture before going deep?
3. Component deep dive (20 minutes): Database choice (SQL vs. NoSQL and why), caching strategy (Redis/Memcached, TTL), API design, handling failures.
4. Trade-off articulation: “What did you sacrifice by choosing eventual consistency here?” L5 candidates who cannot articulate trade-offs — who instead describe a system as if every choice is obviously correct — typically receive “hire with hesitation” or “no hire.”

For L6 loops, the prompt is more ambiguous by design. An L6 candidate is expected to scope the problem themselves, challenge assumptions, and drive the discussion rather than respond to interviewer prompts.

Level and compensation context

Amazon levels software engineers on the same 1–10 scale used company-wide, but SDEs span L4 through L8+. The levels you will encounter when interviewing:

Amazon’s L5 median total compensation of $271K trails Google L5 at $423K and Meta E5 at $480K. The gap is real and worth understanding before negotiating. Amazon compensates partly through Year 1 and Year 2 sign-on bonuses (which front-load cash while RSUs vest on a back-weighted schedule: 5%/15%/40%/40% over four years). If you have a competing offer, Amazon’s compensation team has more flexibility than most candidates realize — a competing offer letter is the most reliable way to move a number.

Six-week prep plan

The candidates who perform best at Amazon treat LP prep with the same rigor as algorithm prep. Six weeks is enough time if you are disciplined.

Weeks 1–2: Build your story bank

Identify five to six real experiences from your career — shipped features, production incidents, disagreements with a manager, cross-functional projects, technical decisions you owned. For each, write a one-page STAR outline. Map each story to at least two LPs. Cover these five LPs with at least one story each: Ownership, Customer Obsession, Dive Deep, Bias for Action, Deliver Results. These are the most frequently assigned.

Weeks 3–4: Algorithm foundations

Do 40–60 LeetCode problems. Focus on medium difficulty. Prioritize: sliding window, DFS/BFS, DP (1D then 2D), heap problems. Time yourself. Practice talking while coding — narrate your reasoning out loud even when practicing alone.

Week 5: System design

Study the standard building blocks: load balancers, CDNs, SQL vs. NoSQL trade-offs, consistent hashing, message queues (Kafka vs. SQS), rate limiting, caching patterns. Practice designing two or three systems end-to-end in 45 minutes. If you are applying for L5 or above, design a distributed write-heavy system (notification service, order tracking) and a read-heavy cache-dependent system (product catalog, URL shortener).

Week 6: Full mock loops

Do at least two full mock interviews per week — one behavioral, one technical. Record yourself if you can. Listen back for: how long your LP stories run (should be 2–3 minutes before follow-ups), whether you speak your coding reasoning aloud, and whether your system design narration sounds like a practitioner or a textbook. Fix the one biggest weakness before your actual loop.

The Amazon loop rewards candidates who have genuinely internalized the Leadership Principles — not as interview prep, but as a framework for how they make decisions. The interviewers have heard every coached answer. What gets hired is real depth: the specifics of what you actually did, the constraints you actually faced, and the results you actually produced.