Apppublic

kumardatascience/Multi-Source-RAG-AI-System-with-Query-Routing

sourceHugging Faceupdated 3mo agoView on Hugging Face

1likes

App README

🤖 AI Chat Assistant

![CI](https://github.com/kumardatascience/ai-chat-assistant-themis/actions/workflows/ci.yml)

🚀 [Live Demo on Hugging Face Spaces →](https://huggingface.co/spaces/kumardatascience/Multi-Source-RAG-AI-System-with-Query-Routing)

An intelligent AI chatbot that routes queries between direct LLM, document retrieval (RAG), and live web search — built with LlamaIndex Workflows, Gemini, ChromaDB, Tavily, and Chainlit.

✨ Features

—💬 ChatGPT-like UI with streaming responses
—🚦 Intelligent query routing (direct / rag / web / multi)
—📚 Document Q&A with ChromaDB
—🌐 Live web search via Tavily
—🧠 Conversation memory
—🔍 Transparent reasoning steps
—🐳 Dockerized
—✅ CI tested via GitHub Actions

🚀 Quick Start

Prerequisites: Python 3.12+, Gemini API key, Tavily API key

bash

# Clone and set up
git clone https://github.com/kumardatascience/ai-chat-assistant-themis.git
cd ai-chat-assistant-themis
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

# Add your API keys
cp .env.example .env
# Edit .env and fill in GOOGLE_API_KEY and TAVILY_API_KEY

# (Optional) Add documents to data/documents/ then index them
python3 src/app/rag.py

# Run
cd src/app
chainlit run main.py -w

Open http://localhost:8000.

🐳 Docker

bash

docker build -t ai-chat-assistant .
docker run --rm -p 8000:8000 --env-file .env ai-chat-assistant

🧪 Tests

bash

pytest -v

📁 Project Structure

src/app/
├── main.py          # Chainlit entry point
├── llm.py           # Gemini wrapper
├── rag.py           # Document indexing + retrieval
├── web_search.py    # Tavily search
└── router.py        # LlamaIndex Workflow router
data/documents/      # Your PDFs/text files
tests/               # Pytest tests
Dockerfile
.github/workflows/ci.yml

🛠️ Tech Stack

Chainlit · Google Gemini · LlamaIndex Workflows · ChromaDB · Tavily · HuggingFace embeddings (BAAI/bge-small-en-v1.5) · pytest · Docker · GitHub Actions

📝 License

MIT