Amazon India Senior MLE / Applied Scientist Interview Guide
Amazon India hires senior Applied Scientist + SDE-ML roles across recommendation (Amazon.in, Prime Video, Amazon Music), forecasting (supply chain), Alexa, AWS ML services, and ads. The interview tests Amazon LP fluency at every step plus deep ML. This guide covers it.
Amazon India has senior ML hiring across many orgs — retail recommendation, supply chain forecasting (one of the largest forecasting ML operations in the world), Alexa, AWS ML services (SageMaker, Bedrock), and ads. Roles include Applied Scientist (research-leaning) and Senior SDE-ML (engineering-leaning). The interview is structurally distinctive: every round is graded against Amazon's Leadership Principles.
The loop structure (Senior AS / SDE-ML)
Standard Amazon loop: phone screen (technical) → on-site 4-5 rounds (each mixes technical + behavioral graded on LPs).
Round-by-round breakdown
Phone screen. 60 min. Coding + ML fundamentals.
On-site round 1 — Coding + LP. 60 min. LeetCode medium-hard + 1-2 LP questions.
On-site round 2 — ML depth + LP. 60 min. Probability, statistics, ML fundamentals (XGBoost internals, regularisation derivations, calibration) + 1-2 LP questions.
On-site round 3 — ML system design + LP. 60 min. Domain-specific (recommendation, forecasting, search). Plus LP.
On-site round 4 — Bar-raiser. 60 min. Hardest round. Senior interviewer from a different org probes deeply across LP + technical.
What Amazon India weights distinctively
1. LP fluency. Every answer must thread an LP (Customer Obsession, Ownership, Invent and Simplify, etc.). No exceptions. 2. STAR format. Behavioral answers in Situation-Task-Action-Result format. Vague stories fail. 3. Frugality + scale. Amazon builds scalable cheap solutions over elegant expensive ones. 4. Long-term thinking. Multi-year impact valued.
Top 10 questions
1. "Design Amazon's product recommendation. Latency, scale, cold-start." 2. "Forecasting: design Amazon's supply chain demand model. SKU × node × day." 3. "Tell me about a time you had to disagree and commit." 4. "Tell me about a time you delivered something with insufficient data." 5. "Calibration in ranking — when does Amazon care?" 6. "Tell me about a time you simplified a complex system." 7. "Design Alexa wake-word detection at low-power edge." 8. "Tell me about a time you had to make a decision quickly under uncertainty." 9. "Real-time feature engineering for product search — architecture." 10. "What's your most ambitious project and why did it matter?"
Prep path through MSL — All Tier 5, Tier 7, Tier 8 posts. Posts 88 (Time Series), 70-72 (RecSys), 125 (Hierarchical Forecasting). Practice: MockInterview + 5-10 LP stories prepped in STAR format.
Common failure modes — Weak LP threading, vague STAR stories, surface ML answers without Amazon scale awareness.
Compensation. Amazon India Senior Applied Scientist / SDE-ML L6 in 2026 ranges ₹50 lakh – ₹85 lakh for 5-8 YOE base + RSU. L7 (Principal) ₹1.2 crore+.