metadata
title: Multi Agent Chat
emoji: π€
colorFrom: blue
colorTo: indigo
sdk: docker
pinned: false
app_port: 7860
π€ Multi-Agent AI System
Production-ready AI backend (FastAPI + LangGraph) with a modern React.js chat frontend.
Try on Huggingface Space
API SwaggerUI
Features
- React Frontend: Gradient UI, chat memory
- Four AI Agents: Weather, Documents (RAG), Meetings, SQL
- Vector Store RAG: ChromaDB semantic search
- Reliable Tool Execution: Deterministic tool calls
- File Upload: PDF, TXT, MD, DOCX support
- One-Command Start:
start.batorstart.sh
Quick Start
Windows:
./start.bat
Linux/Mac:
chmod +x start.sh && ./start.sh
Frontend: http://localhost:3000 Backend: http://localhost:7860
Manual Setup
Backend:
uvicorn main:app --reload
Frontend:
cd frontend
npm install
npm start
Usage Examples
- Weather: "What's the weather in Chennai?"
- Documents: Upload PDF β Ask "What is the policy?"
- Meetings: "Schedule team meeting tomorrow at 2pm"
- Database: "Show all meetings scheduled tomorrow"
Architecture
React UI (3000) β FastAPI (7860) β LangGraph
β
ββββββββββββ¬βββββββββ¬ββββββββββ¬βββββββββ
β Weather β Docs β Meeting β SQL β
β Agent β +RAG β Agent β Agent β
ββββββββββββ΄βββββββββ΄ββββββββββ΄βββββββββ
Configuration (.env)
GITHUB_TOKEN=ghp_... # Optional (GitHub search)
OPENWEATHERMAP_API_KEY=... # Required for weather
Get tokens:
Project Structure
cr-agent/
βββ agents.py # AI agents
βββ main.py # FastAPI server
βββ tools.py # Tool implementations
βββ vector_store.py # ChromaDB RAG
βββ start.bat # One-command startup
βββ frontend/ # React UI
βββ src/App.js
βββ package.json
Documentation
- COMPLETE_SETUP.md: Full setup guide
- FRONTEND_SETUP.md: Frontend details
- TOOL_CALLING_ISSUE.md: Technical analysis
Test Results
- Weather Agent: β Working
- Document RAG: β Working (similarity: 0.59-0.70)
- SQL Agent: β Working
- Meeting Agent: β Working
Tech Stack
- FastAPI, LangGraph, ChromaDB
- React 18, Axios
- sentence-transformers
- Docling
Made with β€οΈ using FastAPI, LangGraph, React, and ChromaDB