VenuMatch
Agentic venue matching system using LangGraph, ChromaDB, FastAPI, and React/Vite.
VenuMatch is an agentic venue matching system built end-to-end in approximately 5 hours as a rapid prototype demonstrating LLM-powered retrieval and reasoning over structured venue data.
Architecture
- LangGraph for agentic workflow orchestration — multi-step query decomposition, retrieval, and answer synthesis
- ChromaDB for vector-based venue embeddings and semantic search
- FastAPI backend exposing search and recommendation endpoints
- React + Vite frontend for user-facing venue discovery interface
Deployment
- Backend deployed via Railway
- Frontend deployed via Vercel
Stack
Python · LangGraph · ChromaDB · FastAPI · React · Vite · Railway · Vercel
Placeholder — add UI screenshot here.