The unofficial study hall for Amazon’s ML Summer School

What it does

This repository is a shared study folder for the Amazon ML Summer School, an Indian campus recruiting program aimed at students graduating in 2027 and 2028. It stockpiles previous years’ MCQs, programming questions, official sample tests, and module summaries covering an eight-week curriculum that runs from supervised learning to reinforcement learning. It is not software; it is a crowd-sourced prep book maintained by someone who survived the program.

The interesting bit

The repo goes beyond simple file-dumping. It includes solved sample tests, distinct question sets from the 2025 morning and afternoon exam shifts, and module-wise coding exercises. The maintainer even notes—with refreshing candor—that other platforms frequently scrape these materials without credit, which inadvertently confirms this is the primary source.

Key highlights

• Past papers from 2023, 2024, and 2025, including two separate sets from the 2025 morning and afternoon shifts
• Eight module summaries spanning supervised learning, deep neural networks, causal inference, and reinforcement learning
• Official 2025 sample test included with solutions
• Extra ML resources section with roadmaps and conceptual notes
• Mirrors the actual Unstop assessment format: 20 MCQs and 2 programming questions in 60 minutes

Caveats

• The 2026 branding is mostly forward-looking; the bulk of the content is drawn from the 2025 cohort and earlier
• PDFs occasionally fail to render in-browser on GitHub, so local downloads are sometimes required
• This is a curated archive of notes and past papers, not a framework or tool

Verdict

Bookmark it if you are an Indian undergrad grinding for the Amazon ML Summer School assessment or internship loop. Everyone else can safely skip it unless they miss the adrenaline of campus recruiting season.