5000 real amazon interview questions (2026): What to Expect in Amazon Recruiter 30-Minute Prescreen Call After OA. Verified candidate reports — coding, system design, behavioral.

亚麻OA后recruiter 30电话面试问什么

Amazon Robotics Intern面经

Amazon亚麻 SDE NG四轮面经

9.8 亚麻intern oa

2026年6-9月 Amazon 面经整理：OA 先变天了

Amazon VO 一轮ood 面筋

香蕉 L6 店面

Amazon 新鲜OA

Amazon OA

亚麻OA

Amazon OA 面经｜Coding + Django Debug

SDE 2 at Amazon

香蕉厂senior as 店面

香蕉厂OA

Amazon AGI AS 全套面经

&#128021; 家R1 挂经

亚麻店面烙印

亚麻店面&#10133;onsite

## Pipeline

A typical Amazon loop runs:

1. **OA** (when applicable, mostly NG / intern / early-career). Two parts: a classic LC-style problem + an AI Assistant Repo Bug-Fix in a multi-file harness. ~70 min total.
2. **Recruiter screen** — process intro + standard BQ, light technical check on title fit.
3. **Phone screen — coding** (60 min). Usually one LC-medium plus an LP-flavored BQ at the front or back.
4. **Onsite / virtual loop** — 4-5 rounds:
   - 2× coding (one of which is often OOD with a multi-step build),
   - 1× system design (or ML system design for AS / MLE),
   - 1× hiring manager BQ deep-dive,
   - 1× Bar Raiser — runs independently of the team, gates the offer.

FAR / Premium loops (frontier-research orgs, Applied Science staff bands) add: **Paper Read Round**, **Science Application (Ambiguous Problem)**, and a **Deep Dive** on prior work.

## Typical timeline

- OA → recruiter → phone screen: 2-4 weeks.
- Phone screen → onsite: 2-4 weeks.
- Onsite → decision: 1-2 weeks (Bar Raiser write-up can extend this).
- Offer → team-confirm: 1-3 weeks.

## Signaling notes

- Bar Raiser write-ups dominate the HC discussion; nailing one LP deep with a quantified outcome can rescue a borderline technical round.
- Several reports describe "strong technical, weak LP" outcomes — Amazon down-levels rather than rejects in this case.
- "Internal references requested early" is **not** a borderline signal at Amazon (unlike at OpenAI) — recruiters routinely ask.
- 2026 NG loops increasingly run as **team-based hiring (“组招”)**: every interviewer in the loop comes from the same hiring team, so the questions and difficulty track that team rather than a fixed company bank — two candidates in the same week can see very different coding bars.
- Intern / NG loops are sometimes compressed: a single screen (most-challenging-project + one coding problem) or two back-to-back 60-minute interviews, increasingly held **in person** at an Amazon office with a 15-min resume/AI-grill segment bolted onto each technical round.

- MLE loops can apply a six-month cooldown after rejection, while separate SDE hiring teams may still be able to proceed independently. Treat cooldown scope as role/team-specific and confirm it with the recruiter.

## Official format anchors

- The SDE II online assessment is officially structured as 118 minutes total: a 90-minute two-question coding block, a 20-minute systems-design scenario block, and an 8-minute Work Style Survey.
- The standard SDE II onsite is officially four 55-minute interviews. Senior bands and FAR / Premium loops layer additional rounds on top of this baseline.
- The phone screen runs in Amazon's internal Livecode editor — syntax highlighting only, code does not execute. Plan to dry-run examples by hand.

## Mid-2026 updates

- NG loops are now frequently **fully in person**: four back-to-back rounds over roughly four hours, most rounds splitting ~30 minutes technical / ~30 minutes LP with a pure-BQ final round, and coding done on a whiteboard.
- L5 team-based phone screens can add an unannounced, team-scoped mini system-design after the coding problem — even when the recruiter frames the round as coding-only.
- When a recruiter moves an application to a different Req / Job ID, a completed OA does not always carry over; a repeat OA on the new requisition is possible.

## Boomerang Loop Flexibility (2026 Q3)

Returning L5 candidates can see a compressed two-interview loop: roughly 20 minutes of Leadership Principles plus 30 minutes of coding, followed by another 20-minute BQ segment and 30 minutes of design. A team-fit follow-up may then become a project-architecture whiteboard and hiring-manager conversation rather than a full restart. When a competing offer deadline is documented, the team can add that conversation quickly and return a decision within hours.

## August 2026 process signal

- An NG loop may split three back-to-back virtual rounds from a separately scheduled Bar Raiser several weeks later; a long gap before the final round is not itself a rejection signal.
- A post-VO SDE II rejection can now carry a one-year cooldown rather than the historically common six months. Cooldown scope can vary by role and hiring team, so confirm the exact end date and whether another requisition is eligible with the recruiter.

## Front-End NG loop variant (August 2026)

Front-end / full-stack new-grad hiring can use a phone screen followed by two virtual onsite rounds and a separately scheduled Bar Raiser. The technical mix is front-end UI implementation, JavaScript/browser fundamentals, one general data-structure problem, and LP coverage throughout; the Bar Raiser remains pure behavioral.

## Late-summer 2026 scheduling variants

NG onsite scheduling varies by team: Vancouver can spread four rounds across two days, while Boston can run four back-to-back interviews. One Boston timeline ran from July 31 OA to August 24 availability survey, September 4 onsite, and September 10 verbal offer; this is an example timeline rather than a response-time guarantee.

A Summer 2026 intern path matched the candidate to a team before interviewing and used two one-hour interviews after the OA, with the hiring manager and an L5 SWE. Both mixed coding and behavioral questions; the manager emphasized behavior and GenAI experience. The eventual offer was for the interviewing team.

For SDE II, a filled requisition can interrupt progress after the OA and lead the recruiter to offer other openings. An onsite availability survey may require a response within 24 hours, while the detailed schedule can arrive only the day before the interviews.

## September 2026 interview-format variants

An Applied Scientist one-hour phone screen can begin with roughly 30 minutes of Leadership Principles before moving to a project deep dive and coding, even when the recruiter describes a half-ML, half-coding agenda without behavioral questions. Prepare behavioral examples alongside the technical material and leave room for changes to the advertised split.

An SDE II phone screen can put technical work first and LP questions last, with an initial coding exercise followed by two separate design discussions. Finishing the first problem quickly does not necessarily end the technical portion.

A four-round virtual loop can also separate a pure behavioral round and a pure coding round, then combine behavior with system design and coding in later rounds. The presence and placement of LP questions varies across individual interview schedules.

## Additional September 2026 team formats

An AI infrastructure hiring path can use a phone-screen object-design exercise followed by five onsite interviews: three coding rounds, one ML round and one system-design round. This differs from the four-round general SDE format; check the team-specific schedule when planning preparation.

An SDE II phone screen can allocate the first 20 minutes to a resume deep dive plus two LP questions, then 40 minutes to coding. Technology choices, trade-offs, business logic and personal responsibility are part of that opening discussion. Continued explanation during coding can consume substantial time: one candidate spent about half of the coding allocation discussing the approach and finished without time for a dry run.


Loop Structure & Timeline

Leadership Principles & Bar Raiser

GenAI BQ — The Universal New Question

Coding Round Style — OOD & Under-Specified

Tooling, Platforms & Logistics

FAR & Premium-Tier Loops (Frontier AI Research / Applied Scientist Staff)

Amazon's bar leans on Leadership Principles and scenario thinking more than raw algorithm difficulty. Coding rounds skew toward OOD and under-specified prompts that reward clarification skill, and a GenAI-usage behavioral question appears in nearly every modern loop. Bar Raiser independence and writing-heavy work samples shape the senior-level signal.

AI Assistant Repo Bug-Fix (OA Part 2)

amazon interview questions (2026)

Real Interview Questionsfrom Real Candidates

Real Interview Questions
from Real Candidates