🥗 RAG Fundamentals Workshop

Welcome to the RAG Fundamentals Workshop by learnwithparam.com!

This hands-on workshop teaches you the core concepts of Retrieval-Augmented Generation (RAG) through a series of progressive, scenario-based notebooks. We'll be building an AI-powered assistant for "Green Bites", a fictional 100% plant-based restaurant.

Regional pricing is available for eligible learners, with discounts of up to 60% in supported regions. Start here: https://www.learnwithparam.com/ai-bootcamp

🎯 What You'll Learn

• Embeddings: How AI represents meaning in numbers.
• Vector Stores: Storing and retrieving knowledge with Qdrant.
• Chunking: Strategies for splitting long documents for better accuracy.
• RAG Pipelines: Building end-to-end question-answering systems.
• Agentic RAG: Using tools and web search (ddgs) for intelligent responses.

---

📚 Workshop Structure

0. 00-introduction.ipynb: Understand the limitations of LLMs and why Retrieval-Augmented Generation (RAG) is the industry-standard solution.
1. 01-embeddings.ipynb: Learn how AI "understands" text by converting the Green Bites menu into vectors.
2. 02-vector-stores.ipynb: Build a searchable knowledge base using Qdrant for restaurant policies.
3. 03-chunking-strategies.ipynb: Compare recursive and semantic splitting on long recipe documents.
4. 04-rag-pipeline.ipynb: Assemble the full RAG pipeline to power the "Green Bites Assistant".
5. 05-agentic-rag-and-tools.ipynb: Give your assistant superpowers with intelligent routing and web search.

---

🛠️ Technology Stack

• LLM Abstraction: LiteLLM (Supports Gemini, OpenAI, etc.)
• RAG Framework: LlamaIndex
• Vector Database: Qdrant
• Search Tool: ddgs (Free & Open Source)
• Environment: UV, JupyterLab, Docker

---

🚀 Getting Started

Prerequisites

• Docker (to run Qdrant)
• UV (for lightning-fast Python setup)
• An AI API Key (Gemini is recommended for its free tier)

Setup

1. Initialize Environment:

   make setup

2. Configure API Keys: Edit the .env file and add your GOOGLE_API_KEY (or OPENAI_API_KEY).

3. Start the Workshop:

   make dev

   This will start Qdrant and launch JupyterLab at http://localhost:8888.

---

📖 How to Use This Workshop

Each notebook follows a Scenario → Problem → Solution structure:

• Scenario: A real-world business need for Green Bites.
• The Problem: Why a simple LLM prompt isn't enough.
• The Solution: Implementing a specific RAG concept to solve it.
• Discussion Questions: Deepen your understanding.
• Challenge Tasks: Practice what you've learned.

---

🎓 About This Workshop

Created with ❤️ by learnwithparam.com.

Happy Coding! 🚀