🧠 GenAI Study Notes¶

Comprehensive, structured study notes for Generative AI — from prerequisites to frontier research. 83 curriculum-ordered notes. Career-mapped to 31 AI/ML roles. Research based Human & AI created, reviewed and updated.

📖 Browse the Docs Site →

📑 Table of Contents¶

🧠 GenAI Study Notes
📑 Table of Contents
What This Is
⚡ Quick Start
🗺️ How to Use
📚 What's Covered
🛠️ Study Tools
📝 Note Format
Quality Notice
🤝 Contributing
⚖️ License

What This Is¶

This repository is a public collection of structured GenAI study notes in plain markdown, published as a full-featured docs site at bhaskarjha-com.github.io/genai-notes. Each topic is designed as a self-contained 30–60 minute study session that combines theory, code, terminology, comparisons, gotchas, and interview angles.

The repo is organized by learning sequence rather than alphabetically. You can browse the docs site for the best reading experience, open the repo as an Obsidian vault for graph-based exploration, study linearly via the Learning Path, jump in by role, or use individual notes as a portable reference in GitHub or any markdown editor.

⚡ Quick Start¶

git clone https://github.com/bhaskarjha-com/genai-notes.git
cd genai-notes

Then pick your path:

Path	Start Here	Best For
Docs site	bhaskarjha-com.github.io/genai-notes	Best reading experience — search, nav, dark mode
Linear study	LEARNING_PATH.md	Systematic coverage, universal foundation + role tracks
Role-based study	career/roles/	Jump to your target role's curated learning path
Topic reference	genai.md	Browse the full scope map and dive into any topic
Obsidian vault	Open this folder in Obsidian	Graph view, backlinks, local-first study

🗺️ How to Use¶

Path 1: Linear Study¶

Start with LEARNING_PATH.md. It contains the universal foundation plus 5 role-cluster tracks.

Path 2: Role-Based Study¶

Read career/genai-career-roles-universal.md, then jump into one of the dedicated role guides:

If your target role is more specialized or niche, use the grouped guides linked from career/genai-career-roles-universal.md alongside the universal role reference.

Path 3: Topic Reference¶

Use genai.md as the root GenAI overview note, then browse the folders below for deeper dives.

Path 4: Obsidian Vault¶

This repo works as an Obsidian vault out of the box — pre-configured with graph colors and excluded folders.

Setup (one-time): 1. Install Obsidian (free) 2. Open Obsidian → "Open folder as vault" → select this repo's root folder 3. That's it — the committed .obsidian/ config handles the rest

What you get: - Graph View (Ctrl+G) — 83 notes + 300 connections, color-coded by domain - Backlinks — click any note to see every note that links to it - Quick Open (Ctrl+O) — type any topic name to jump instantly - Local-first — works offline, your data stays on your machine

Note: If you create or edit notes in Obsidian, always use standard markdown links ([text](path.md)), not wiki-links ([[text]]). The CI pipeline will reject wiki-links. The included .obsidian/app.json config enforces this by default.

📚 What's Covered¶

Current snapshot: 83 topic notes across 13 knowledge directories, a root genai.md overview, dedicated and grouped career guidance, generated learner tooling, and reusable templates.

genai-notes/
├── prerequisites/        (7 notes)   math, deep learning, and NLP prerequisites
├── foundations/          (9 notes)   transformer and representation fundamentals
├── llms/                 (5 notes)   model behavior, selection, and failure modes
├── techniques/           (16 notes)  prompting, RAG, fine-tuning, merging, and adaptation
├── applications/         (8 notes)   applied product and interface patterns
├── tools-and-infra/      (6 notes)   tooling, data, cloud, and systems foundations
├── production/           (14 notes)  serving, ops, CI/CD, reliability, and system design
├── evaluation/           (5 notes)   benchmarks, deep-dive evaluation, interview prep
├── inference/            (4 notes)   optimization, GPUs, and distributed serving
├── ethics-and-safety/    (7 notes)   governance, regulation, and security
├── multimodal/           (4 notes)   multimodal and computer-vision fundamentals
├── research-frontiers/   (4 notes)   interpretability, training, and research practice
├── agents/               (6 notes)   AI agents, multi-agent, protocols, and evaluation
├── career/                           universal role map, dedicated + grouped guides
├── learner-tools/                    docs-site learner tooling pages
├── downloads/                        generated decks and progress checklists
├── generated/                        generated markdown reference outputs
├── assets/                           site CSS, JS, and generated JSON data
├── _templates/                       note and career-guide templates
├── genai.md                          root GenAI overview note
└── root docs                         README, LEARNING_PATH, CONTRIBUTING, CHANGELOG, LICENSE

🛠️ Study Tools¶

Anki-friendly interview decks
Progress tracker downloads
Topic × role relevance matrix
Interactive docs-site tooling, including the knowledge graph, role matrix, and browser-based progress tracker

📝 Note Format¶

Notes follow a consistent template so the structure stays predictable.

TL;DR gives the fast overview.
Overview and Deep Dive cover the core knowledge.
Connections links the note to the rest of the repo.
Sources keeps the trail back to original material.

The template uses section markers:

★ sections are the core sections every note must have.
◆ sections are included when the topic calls for them.
○ sections are optional but useful when relevant.

Quality Notice¶

These notes are created, reviewed and updated by human & AI. We aim for accuracy, but advanced topics can still benefit from expert review. If you spot an issue, please open an issue or pull request.

🤝 Contributing¶

See CONTRIBUTING.md.

⚖️ License¶

This repository is licensed under CC-BY-4.0.