Docs / Interview Guides / Amazon Interview Guide

Amazon Interview Guide

Amazon's loop is the most behavioral-heavy of the FAANG-tier interviews. The Leadership Principles aren't a marketing line — they're scored explicitly, often outweighing pure coding ability. Engineers who underprep behavioral questions reliably fail this loop.

Compensation (US, approximate)

LevelBase + Bonus + Stock (4-yr avg)
SDE I (L4)$170K – $200K
SDE II (L5)$230K – $310K
Sr. SDE (L6)$360K – $480K
Principal (L7)$550K+

Amazon's stock vest is heavily back-loaded (5/15/40/40 over four years), which inflates first-year offers but also creates pressure to stay through year four.

Interview Process

  1. Recruiter Phone Screen (20–30 min) — Resume, team fit, level signal. No coding.

  2. Online Assessment (1–2 rounds) — Algorithmic coding problems plus a work-style assessment. The OAs are auto-graded; passing is binary.

  3. Technical Phone Screen (45–60 min) — One coding problem on a shared editor. Expect data-structures-and-algorithms questions plus 1–2 behavioral questions.

  4. Onsite Loop (5–6 interviews, ~5 hours)

    • Coding rounds (2–3) — solved on a whiteboard or virtual whiteboard, medium-to-hard difficulty.
    • Design (1 round) — OOP design for SDE I/II, system design for SDE II+ and Senior.
    • Bar Raiser (1 round) — a trained, neutral interviewer from a different team whose job is to enforce hiring quality. Their vote can override the hiring manager.
    • Behavioral — every round opens or closes with 1–2 behavioral questions tied to specific Leadership Principles.

What Amazon Actually Evaluates

Amazon scores candidates against the 16 Leadership Principles, with each interviewer assigned 2–3 principles to probe. The technical bar matters, but a candidate who codes well and bombs behavioral will not get an offer. Conversely, average-but-clear coding plus strong behavioral often does.

The most-probed principles for software engineers:

  • Customer Obsession — every decision should trace back to user impact.
  • Ownership — you take responsibility beyond your immediate scope.
  • Bias for Action — you ship, you don't analyze forever.
  • Deliver Results — outcomes matter more than effort.
  • Earn Trust — you give and receive feedback honestly.
  • Dive Deep — you can go to first principles when needed.

Interview Tips

  1. Prepare 12–15 STAR stories. Situation, Task, Action, Result. Each story should map to 2–3 Leadership Principles. Practice telling them in 2–3 minutes max.

  2. Focus on "I" not "we". Amazon explicitly scores individual contribution. "We launched X" gets followed up with "what did you do?"

  3. Bring data. "I improved latency" is weaker than "I cut p99 latency from 800ms to 120ms by replacing the synchronous fanout with a Kafka topic, which unblocked the recommendation team's launch."

  4. Code clearly, but don't over-optimize. Amazon's coding bar is medium — the bar raiser is checking for clean thinking, not a perfect solution.

  5. For senior roles, expect ambiguity in design. Amazon's design questions are often vague on purpose ("design a logistics system"). Asking clarifying questions is part of the score.

Topics That Come Up Often

  • Graphs — BFS, DFS, topological sort, shortest path.
  • Trees — traversals, LCA, serialization.
  • Strings & Arrays — sliding window, two pointers.
  • Design — distributed system primitives, queue-based pipelines, sharded data stores.

Reference