RAG Chatbot

A simple Retrieval-Augmented Generation (RAG) chatbot that can answer questions based on your own documents. Can be use internally or to helps your clients to use your application.

Built with FastAPI, FAISS, and Gradio for a lightweight local assistant setup.

---

How to Run

1. Create a virtual environment

python3 -m venv venv
source venv/bin/activate

2. Install dependencies

pip3 install -r requirements.txt

3. Run the app

./run.sh

---

Description

This chatbot:

• Uses a RAG pipeline to combine LLM reasoning with document retrieval.
• Stores embeddings locally in a FAISS index.
• Exposes a simple FastAPI backend and a Gradio chat interface.

---

Stack

• Python 3.11+
• FastAPI
• FAISS
• Sentence Transformers
• OpenAI / compatible LLM
• Gradio

License Under MIT license, see LICENSE.