What it is

genieincodebottle/generative-ai is a public GitHub repository that gathers Generative AI learning materials in one place: PDF guides, Jupyter notebooks, cloud-setup documents, interview prep, and ready-to-run example projects (RAG, agents, chatbots, etc.). It positions itself as a “go-to hub for end-to-end GenAI learning.”

What you'll find inside

Learning & documentation

• Core concepts: PDFs on vector embeddings, a Jupyter notebook on prompt engineering, AI design patterns, and a machine learning reference guide.
• Architecture: PDFs covering tech stacks, comparisons of LLM providers, advanced RAG decision flows, and a GenAI project lifecycle.
• Cloud platforms: Guides for deploying GenAI on AWS, Azure, and Google Vertex AI.
• Interview prep: Q&A modules and PDFs for GenAI, RAG/prompting, LLMs, Agentic AI, ML system design, MLOps, and more.
• Production: A checklist for enterprise production-readiness.

Hands-on projects (in the genai-usecases folder)

• Retrieval-Augmented Generation (RAG): Advanced RAG with 9 patterns (e.g., corrective RAG, hybrid search, query expansion) plus a cache-augmented alternative.
• Agentic AI: Multi-agent systems using CrewAI and LangGraph, 23 reasoning patterns (Chain-of-Thought, ReAct, etc.), a content moderation system, and a Model Context Protocol (MCP) example.
• Conversational AI: A PDF chatbot with persistent memory; a full-stack customer-feedback analytics app.
• LLM tools: Comparisons of OpenAI, Gemini, Claude, Groq, and local models; embedding model guides.
• Data & analytics: Text-to-SQL, natural-language querying of Neo4j graphs, sentiment analysis of call transcripts, and a chat analytics dashboard.
• Prompt security: 16+ prompt engineering techniques; a prompt-injection detector using Meta's Llama Guard.
• Multimodal: A text-to-image app (Gemini 2.5 Flash) and an all-in-one Llama 4 app (chat, OCR, RAG, agents).
• Automation: A setup guide for n8n workflow automation.

Related external projects

Several separate open-source tools are listed as recommended resources (not code in this repo), including local coding agents (Tura), QA agents (agent-qa), and CLI orchestrators (YYLO, agent-manager).

Companion learning site

A linked platform, AI-ML Companion (aimlcompanion.ai), offers an interactive version of much of this material: 28 tracks, 400+ modules, quizzes, visualizations, and 23 real-world projects, described as “free to start.” The repo's guides are described as the continuously-updated reference material for this site.

How to contribute

Anyone can contribute by forking the repo, cloning it, creating a feature branch, committing changes, pushing, and opening a pull request with a short description.