A production-ready Retrieval Augmented Generation (RAG) system built for the Ready Tensor Agentic AI Developer Certification Module 1. This project demonstrates core concepts learned in Weeks 1-3, including document processing, vector embeddings, semantic search, and LLM integration.
What It Does
The RAG system:
Loads Documents: Ingests multiple text and markdown files
Processes Content: Intelligently splits documents into semantic chunks
Creates Embeddings: Converts text to vector representations for semantic search
Answers Questions: Uses an LLM to generate answers grounded in document content
Attributes Sources: Shows which parts of documents were used to answer questions
Maintains Context: Remembers conversation history across multiple turns
Technologies Used
LangChain 1.0+: Framework for building LLM applications
Groq API: Fast LLM inference with llama-3.3-70b-versatile
HuggingFace Embeddings: Semantic text embedding model (all-MiniLM-L6-v2)
Chroma: Vector database for storing and retrieving embeddings
This RAG system demonstrates a complete, production-ready implementation of Agentic AI concepts from Module 1. It showcases proper software engineering practices while delivering real functionality for document-based question answering.