GenAI Foundry

📢 About GenAI Foundry

This repository was created to make the world of Generative AI accessible, simple, and hands-on.

Our mission:

• 🚀 Empower learners to experiment with LLMs and RAG pipelines.
• 🧠 Provide clear, beginner-friendly tutorials.
• 📚 Foster creativity and exploration with GenAI tools.

Feel free to fork, adapt, and expand these examples! 🎯

📚 Tutorials and Notebooks

Welcome to the GenAI Foundry learning series! Below you'll find a list of hands-on tutorials designed to help you learn about Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), and more.

A Practical Tour of LLM Mechanics (No API Key Required)

📓 Notebook	📝 Description	🚀 Colab Link
1. Tokens and Embeddings	Introduction to tokens and embeddings.
2. Semantic Similarity	Introduction to semantic similarity.
3. Search with Embeddings (RAG Foundations)	Embeddings and RAG.	🚧 Coming Soon
4. Model and Tradeoffs	Model selection and tradeoffs.	🚧 Coming Soon

---

Inroductory Concepts (OpenAI API Key Required )

📓 Notebook	📝 Description	🚀 Colab Link
Simple Chatbot	Chatbot using an API key.
Temperature and Token Explorer	Explore how temperature and tokens affects LLM output.
Prompt Engineering Basics	A gentle introduction to prompt engineering basics.
Few Shot / Zero Shot prompting	Prompt engineering techniques.
Simple RAG Application	A gentle introduction to RAG.
LangChain Basics	A gentle introduction to LangChain capabilities.
Fine-Tuning Basics	Intro to fine-tuning foundation models.	🚧 Coming Soon

---

🎓 Applied Generative AI — Interactive Demos

A multi-page Streamlit application with three interactive demos, all running in a single Docker container.

🔍 RAG Visual Explorer

No API key required

See how Retrieval-Augmented Generation works under the hood:

• Chunking — Watch documents split into searchable pieces
• Embeddings — See text transform into numerical vectors
• Vector Space — Visualize how similar content clusters together
• Semantic Search — Compare keyword vs. meaning-based search

🧠 Ontology & Counterfactual Reasoning

OpenAI API key required for AI features (Demo Mode works without)

Explore structured knowledge and "what if" analysis:

• Ontology Explorer — Interactive knowledge graph of cybersecurity risks
• Counterfactual Analysis — Test what happens when controls fail
• LLM Validation — Verify AI outputs against ground truth
• Concepts — Theory explained simply

🤖 RAG Chat with Estel

OpenAI API key required

A complete document-grounded chat assistant powered by RAG:

• Upload — Load your own PDF or TXT documents
• Index — Documents are chunked, embedded, and stored in a FAISS vector database
• Chat — Ask questions and get answers grounded in your uploaded content
• Learn — See RAG in action as a complete end-to-end pipeline

📖 How RAG Chat Works — Deep Dive Guide

---

🐳 Running the Demos (Docker Required)

📖 Detailed Docker Guide for Beginners

Step	Command
1. Clone this repo	git clone https://github.com/dlwhyte/GenAI_foundry.git
2. Navigate to repo	cd GenAI_foundry
3. Build container	docker build -t genai-foundry .
4. Run demos	docker run -p 8501:8501 genai-foundry
5. Open browser	http://localhost:8501

With an OpenAI API key (required for Ontology demo + RAG Chat):

docker run -p 8501:8501 -e OPENAI_API_KEY=sk-your-key-here genai-foundry

💡 Without the -e flag, you can still enter your API key in the sidebar on any page that requires it.

🔑 How to Get an OpenAI API Key

💡 Don't have Docker? Download it from docker.com/products/docker-desktop

---

🛠 How to Get Started

Action	Description
	✅ New to GitHub? Start here.
	✅ Learn how to install and use Docker for running demos.
	✅ Open and run the notebooks directly in Google Colab.
	✅ Create your OpenAI API key to enable model access. Step-by-step instructions.

---

📝 License

This project is licensed under the MIT License — free to use, modify, and share!

See the LICENSE file for more details.

---