The Amazon interview loop — what to expect
Amazon's SDE interview has four stages. First, the Online Assessment (OA) on HackerRank: two coding problems in 70–90 minutes. These are typically Medium difficulty — array/string manipulation or greedy problems. No debugging step; you pass or fail based on test cases. Second, a Technical Phone Screen (30–45 min): one coding problem on Amazon's internal CodePair tool. Third, the Virtual On-Site — 4 rounds for SDE I, 5 for SDE II+. Each round lasts 55 minutes. Rounds cover: coding (2 rounds), system design (SDE II+, 1 round), and behavioral with LP questions woven throughout every round. Fourth, the Bar Raiser — a special interviewer (not from the hiring team) in one on-site slot whose job is to assess whether you raise Amazon's hiring bar. The Bar Raiser can veto any candidate the loop otherwise wants to hire.
DSA preparation — what Amazon actually asks
Amazon's coding problems are a mix of Medium and occasionally Hard LeetCode-style problems. Unlike Google (which skews toward tricky algorithmic problems) or Meta (which leans on graph/DP), Amazon tends toward practical problems with a design flavor. Most-asked topics by frequency: 1. **Arrays & Strings** — two pointers, sliding window, sorting, string manipulation 2. **Trees** — BFS level-order traversal, DFS variants, LCA, serialize/deserialize 3. **Heaps** — K-th largest/smallest, top-K, merge K sorted lists 4. **Graphs** — BFS/DFS, connected components, shortest path (Dijkstra/BFS on unweighted) 5. **Dynamic Programming** — knapsack variants, edit distance, longest sequences 6. **Design Problems** — LRU cache, hit counter, rate limiter (these appear in "coding" rounds at SDE II+) Target list: Top K Frequent Elements, LRU Cache, Meeting Rooms II, Word Ladder, Minimum Number of Arrows to Burst Balloons, Number of Islands, Longest Substring Without Repeating Characters, Product of Array Except Self, Merge Intervals, Pacific Atlantic Water Flow.
Online Assessment (OA) strategy
The OA is pass/fail. You need to pass most test cases (90%+ to advance reliably). Strategy: **Before submitting:** Test your solution against edge cases manually — empty input, single element, all duplicates, negative numbers. Amazon's test cases specifically target boundary conditions. **If stuck:** Write a brute-force solution first. A working O(n²) solution that passes 70% of test cases is better than an unfinished O(n log n) attempt that passes 0%. **Time management:** 70 minutes / 2 problems = 35 minutes each. If problem 1 is Hard, don't spend more than 45 minutes — move to problem 2. **Common OA formats:** Two coding problems + a Work Style survey (not graded) + a Work Simulation (short scenario questions — answer as a strong senior engineer would).
System design — the Amazon way
Amazon system design interviews are practical, not academic. The interviewer wants to see that you think like an engineer who ships production systems. Framework: **Step 1 — Requirements (5 min):** Ask about scale (QPS, data volume, SLAs), consistency requirements, and which features are in scope. Amazon interviewers reward explicit scoping. **Step 2 — High-level design (15 min):** Draw the core components: client → API gateway → service → database. Name the AWS services you would use (S3, DynamoDB, SQS, ElastiCache, ECS). **Step 3 — Deep dive (25 min):** The interviewer will guide you. Common deep dives: database schema + indexing, caching strategy, handling failures (retries, circuit breakers), and scaling (partitioning, read replicas). **Amazon-specific themes:** Durability over availability (Amazon will lose money if data is lost), eventual consistency is acceptable for non-financial data, S3 and DynamoDB come up constantly as solutions.
The 16 Leadership Principles — complete prep guide
Amazon evaluates behavioral fit through 16 Leadership Principles (LPs). Every interviewer is assigned 2–3 LPs to assess in their round. Prepare STAR stories for each: **Customer Obsession:** "Leaders start with the customer and work backwards." → Tell a story where you made a decision specifically because it was better for users, even if it was harder for engineering. **Ownership:** "Leaders are owners... they act on behalf of the entire company." → Tell a story where you fixed a problem you weren't assigned to, or prevented future issues proactively. **Invent and Simplify:** "Leaders expect and require innovation." → A story where you found a simpler or novel solution to a complex problem. **Are Right, A Lot:** Good judgment + seeking diverse perspectives. → A story where you gathered data, challenged assumptions, or changed your mind based on evidence. **Learn and Be Curious:** Continuous self-improvement. → A story where you learned something outside your comfort zone and applied it. **Hire and Develop the Best:** Raising the bar in hiring, mentoring. → A story of mentoring, onboarding, or identifying a high-potential colleague. **Insist on the Highest Standards:** Never settle for "good enough." → A story where you caught a quality issue others missed or pushed back on a substandard solution. **Think Big:** Bold directions, big picture. → A story where you proposed or led a direction that had outsized impact. **Bias for Action:** Speed matters, risk-taking is valued. → A story where you made a decision with incomplete information and it paid off. **Frugality:** Accomplish more with less. → A story where you solved a problem with limited resources or avoided unnecessary spend. **Earn Trust:** Transparency, self-criticism. → A story where you admitted a mistake and how you recovered. **Dive Deep:** Leaders stay connected to details. → A story where you dug into data or details to uncover a root cause. **Have Backbone; Disagree and Commit:** Respectful challenge, then full commitment. → A story where you pushed back on a decision, were overruled, and then fully committed to the other direction. **Deliver Results:** Execution focus. → A story of overcoming obstacles to hit a goal on time. **Strive to be Earth's Best Employer:** Creating a positive team environment. → A story of supporting team wellbeing or inclusivity. **Success and Scale Bring Broad Responsibility:** Long-term societal impact. → A story of sustainable or ethical decision-making. **STAR-L format:** Situation → Task → Action → Result → **Learning** (Amazon interviewers often ask "what would you do differently?")
Compensation — what to expect at Amazon in 2026
Amazon's compensation structure is unique: it uses a **restricted stock unit (RSU) vesting cliff** and a **signing bonus** to offset the lower first-year base compared to Google/Meta. Base salary is capped at $350K for most roles (Amazon policy). The rest of total compensation comes from RSU grants and sign-on bonuses. **Typical 2026 ranges (US, Seattle/SF):** - SDE I (0–2 YoE): Base $150K–$180K, TC $200K–$260K - SDE II (3–5 YoE): Base $180K–$220K, TC $280K–$380K - SDE III/Senior (6+ YoE): Base $200K–$250K, TC $380K–$550K+ IMPORTANT: Amazon RSUs vest on an **unusual schedule** — 5% Year 1, 15% Year 2, 40% Year 3, 40% Year 4. This means your Year 1 and Year 2 TC are significantly lower than your Year 3–4 TC. The sign-on bonus (paid in Year 1 and sometimes Year 2) compensates for the lower early vesting.