INTERVIEW GUIDE
Amazon Data Scientist Interview: Questions & Process
The interview process
Questions you're likely to get
Technical
- Write SQL to find each customer's most recent order and the days since their previous one.
- Explain p-value and statistical significance to a product manager with no stats background.
- What's the bias-variance tradeoff, and how does it show up when a model overfits?
- For a fraud-detection model, would you optimize for precision or recall? Defend the choice.
- How would you size an A/B test — what determines the sample size you need?
- How do you detect outliers in a dataset, and when should you keep them rather than drop them?
Role-specific
- Conversion just dropped 10% week-over-week. Walk me through how you investigate the cause.
- Design a metric to measure whether a recommendation feature is actually working.
- You can't run an A/B test for a launch. How do you still estimate its impact?
Behavioral
- Tell me about a time you used data to change a decision a stakeholder had already made (Dive Deep).
- Describe a time you disagreed with your manager and held your position (Have Backbone; Disagree and Commit).
- Tell me about a time you took ownership of an ambiguous, undefined problem (Ownership).
- Describe a time you delivered a result under a tight deadline with imperfect data (Deliver Results).
Practice these problems live
Relevant LeetCode problems for the Amazon Data Scientist loop. Start a live, AI-run coding interview on any of them — or open the problem on LeetCode.
How to answer (worked examples)
What Amazon looks for
- Structured problem-solving — you decompose ambiguous questions instead of guessing
- Statistical rigor — you reason about significance, sample size, and confounders, not just point estimates
- Business sense — you connect metrics to customer and revenue impact, not just model accuracy
- Strong, specific STAR stories mapped to Leadership Principles with quantified results
- Clear communication — you can explain technical findings to a non-technical audience
- Red flag: vague behavioral answers ('we' instead of 'I'), or technical answers with no business framing
FAQ
How important are the Leadership Principles for a Data Scientist role?
Critical. Behavioral LP questions appear in almost every round, and the Bar Raiser weights them heavily. Strong technical skills won't save you if your STAR stories are thin — prepare 6-8 specific stories mapped to LPs like Dive Deep, Ownership, and Have Backbone.
Is there a coding round, or is it mostly SQL?
SQL is the core technical screen, and it can get advanced (window functions, multi-table joins). Some loops add light Python/algorithmic coding, but heavy LeetCode-Hard is less common for DS than for SWE.
How much ML theory do they expect?
Fundamentals you can reason about out loud — bias-variance, regularization, evaluation metrics, A/B testing, common algorithms and when to use them. They care more about sound reasoning than deriving proofs.
What is the Bar Raiser round?
An interviewer from outside the hiring team trained to keep Amazon's hiring bar high. They focus heavily on Leadership Principles and can veto an offer, so treat that round as seriously as the technical ones.
How long is the process?
Typically 3-6 weeks from recruiter screen to decision, depending on scheduling and team availability.
The hardest part isn't the SQL — it's delivering crisp LP stories and structured analysis out loud under a Bar Raiser's questions. Practice this exact interview with OfferLoop's realtime voice coach so the answers come out clean when it counts.
Practice this interview out loud →Related
OfferLoop is an independent interview-practice tool and is not affiliated with, endorsed by, or sponsored by Amazon. All company names and trademarks are the property of their respective owners.
Interview formats vary by team, level and year, and this guide is compiled from general knowledge of publicly discussed hiring processes — treat it as preparation material, not an official description of Amazon's current process.