The Amazon new grad interview process

The stages

The bar raiser is worth knowing about: an experienced interviewer from an unrelated team, trained to hold a consistent standard, whose assessment carries disproportionate weight. You will usually not be told which round it is.

The online assessment

Amazon's is the most distinctive OA among large employers, because it is not only code. Alongside two coding problems there are work-style questions and sometimes a workplace simulation. The coding half is standard - arrays, strings, hash maps, graphs - and the format details are in the new grad online assessment guide, with the Amazon-specific variant in the Amazon online assessment.

• Answer the work-style section consistently, not strategically. It is checking for coherence across items, and trying to game it tends to produce contradictions.
• Finish the coding problems. Passing hidden tests is the whole score; elegance earns nothing.
• Expect it fast. The assessment often arrives within days of applying, which is why preparation should start before you apply.

Leadership Principles, which decide more than people expect

Every behavioral question maps to one or more of the Leadership Principles, and interviewers record which. This is unusually explicit compared with other large employers, and it makes the behavioral half genuinely preparable.

• Prepare six to eight real stories, each usable for two or three different principles. Trying to prepare one per principle is more work than it is worth.
• Use STAR and weight it toward the action. Expect follow-ups drilling into what you specifically did.
• Have failures ready. Several principles are best evidenced by something that went wrong and what you changed, and candidates who only bring successes score worse.
• Quantify where you honestly can, and say so plainly when you cannot.

The full set and worked examples are in Amazon leadership principles examples, and the general structure is in STAR answers that actually land.

The coding half

Difficulty is medium and the patterns are conventional: arrays and hashing, two pointers, trees and graphs, binary search, and entry-level dynamic programming. The Blind 75 covers the range, and the recognition skill is in the 15 patterns guide. Amazon's specific question history is in Amazon coding interview questions.

One honest caveat: Amazon's process has changed repeatedly between hiring seasons, and team-level variation is real. Treat what your recruiter tells you as authoritative over any public description, including this one.

Where CoPilot Interview fits

Disclosure: we build an interview assistant, so treat this as a disclosed pitch. CoPilot Interview is a native Windows and macOS app that puts suggested answers on your own screen in about four seconds, across coding and behavioral rounds, with a permanent free tier that needs no credit card. For a loop weighted this heavily toward prepared stories, the preparation is where most of the gain is.

Frequently asked questions

How many rounds is the Amazon new grad interview?

Typically an online assessment, a technical phone screen, and then a loop of three to four rounds, one of which is the bar raiser. The exact shape varies by team and hiring season, so treat your recruiter's description as authoritative.

What is a bar raiser at Amazon?

An experienced interviewer from a team unrelated to the one hiring, trained to hold a consistent standard across the company, whose assessment carries disproportionate weight. You are usually not told which round it is.

How important are the Leadership Principles for new grads?

Very. Every behavioral question maps to one or more principles and interviewers record which, which makes this half of the loop unusually explicit and unusually preparable. Candidates who prepare only coding tend to underperform here.

How many stories should I prepare for an Amazon interview?

Six to eight real stories, each flexible enough to answer two or three different principles, rather than one story per principle. Include failures, since several principles are best evidenced by something that went wrong and what you changed afterwards.