cu-sanjay/amazon-ml-summer-school-2026

1. This repository is a curated learning archive for the Amazon ML Summer School, not an application. Its “architecture” is primarily informational: a top-level landing page, module-wise lesson notes, assessment/practice artifacts, and a supplemental resource library.

2. The root README.md is the central entry point and navigation hub. It presents the program overview, official application links, assessment format, curriculum timeline, and links out to the main content areas. For diagramming, treat this as the front door of the repository.

3. The AMSS 2025/ directory is the core curriculum subsystem. It contains MODULE1.md through MODULE8.md, each representing one instructional unit in the 2025 syllabus:

- Supervised Learning

- Deep Neural Networks

- Dimensionality Reduction

- Unsupervised Learning

- Probabilistic Graphical Models

- Sequential Learning

- Causal Inference

- Reinforcement Learning

4. Each module file appears to be a self-contained lesson summary and study guide. The README explicitly says these module pages include practice tasks and coding exercises at the end, so the primary content flow is “topic explanation → practice/revision prompts.”

5. The assessment-practice subsystem lives at the root alongside the curriculum. Programming.md provides programming questions for test-style practice, while SummerSchool2023.md captures memory-based previous-year questions for revision. These are separate from the module notes and serve a test-preparation role.

6. The PDF files are archival exam and sample-test artifacts. Amazon_ML_Summer_School_Sample_Test.pdf is the official sample assessment, and AMSS_2025_Exam_Questions.pdf / AMSS_2025_Questions.pdf are real or memory-based exam question collections. They function as reference inputs for self-study rather than as executable assets.

7. More ML Resources/ is an expanded knowledge base and reference library. Its README points to additional notes, roadmaps, and learning material, while the directory also includes major textbook PDFs covering deep learning, causal inference, probabilistic graphical models, reinforcement learning, transformers, and applied ML.

8. The content flow is largely one-way: external program sources and community-collected exam material are curated into markdown summaries and PDFs, then surfaced through the root README for learners. There is no evidence of generated output, data processing pipelines, or stateful runtime behavior.

9. External boundaries matter because several links leave the repository: Unstop hosts the application flow, Amazon/Scaler provide official program pages, and some materials are hosted on GitHub or linked PDFs. These are presentation and sourcing dependencies, not code dependencies.

10. There is no build system, service layer, backend, or deployment target in this repository. The main architectural concern is information organization: a documentation-centric knowledge base designed for browsing, revision, and exam preparation.

11. For an architecture diagram, the cleanest structure is:

- Root landing/documentation layer

- Curriculum module layer

- Practice and previous-question layer

- Extended reference library

- External official links and hosted assets

12. In short, this repo is a learning content hub for AMSS 2025/2026 preparation, with the root README acting as the orchestrator and the rest of the repository serving as modular study, practice, and reference material.