📚 AI-Powered Study Assistant using RAG

This is a project developed as a practical exercise for the Agents course by Hugging Face.

The main goal is to create an AI agent capable of supporting the learning process through Retrieval-Augmented Generation (RAG), leveraging LlamaIndex to ingest, index, and query personal study documents.

---

🎯 Project Goals

• ✅ Practice concepts from the Hugging Face Agents course.
• 📄 Build an intelligent assistant that can answer questions based on personal documents.
• 🤖 Integrate RAG, LlamaIndex, ChromaDB, and Ollama into a seamless workflow.
• 📚 Enhance study effectiveness by turning static notes into an interactive tool.

---

🛠️ Tech Stack

• LlamaIndex – document ingestion, indexing, and retrieval
• ChromaDB – persistent vector store
• Ollama – local LLMs for fast, private responses
• Hugging Face Embeddings (BAAI/bge-small-en-v1.5)
• Python 3.10+

---

🚀 How It Works

1. Documents are loaded from a specified folder.
2. They are split into sentences and converted into embeddings.
3. The embeddings are stored in ChromaDB.
4. You can ask questions in natural language and get context-aware responses powered by a local LLM.