YouQuery

YouTube GraphRAG — Ingest playlists, build a knowledge graph, and query with natural language.

Structure

• backend/ — FastAPI Python API (ingest, query, stats)
• frontend/ — React web UI

Setup

1. Environment

Copy .env.example to .env and fill in your credentials (Neo4j Aura, OpenAI, etc.).

2. Backend

cd backend
python -m venv .venv
source .venv/bin/activate   # or .venv\Scripts\activate on Windows
pip install -e .

Test Neo4j connection:

cd backend && python test_neo4j_connection.py

Run the API:

cd backend && uvicorn main:app --reload --host 0.0.0.0 --port 8000

CLI (optional):

cd backend && python cli.py ingest "https://youtube.com/playlist?list=..."
cd backend && python cli.py query "What did they say about X?" --mode hybrid
cd backend && python cli.py stats

3. Frontend

cd frontend
npm install
npm run dev

Open http://localhost:5173. The dev server proxies /api to the backend.

Usage

1. Ingest — Paste a YouTube playlist URL and click Ingest. Videos are transcribed, chunked, and stored in Neo4j with embeddings.
2. Query — Ask questions in natural language. Choose hybrid (default), semantic, or graph mode.
3. Stats — View playlists, videos, and chunks in the database.

Production

• Backend: uvicorn main:app --host 0.0.0.0 --port 8000
• Frontend: npm run build then serve the dist/ folder. Set VITE_API_URL to your backend URL.