FastAPI + PostgreSQL + pgvector

A minimal FastAPI service demonstrating pgvector usage with async SQLAlchemy.

Features

• GET / — Hello World
• GET /health — Health check
• POST /vectors/ — Store a vector embedding with a label
• GET /vectors/ — List all stored vectors
• POST /vectors/search — Find nearest neighbours by L2 distance

Quick Start

cp .env.example .env
# Edit .env as needed
docker compose up --build

The API will be available at http://localhost:8000 Interactive docs: http://localhost:8000/docs

Environment Variables

Variable	Default	Description
POSTGRES_USER	postgres	DB username
POSTGRES_PASSWORD	postgres	DB password
POSTGRES_DB	vectordb	Database name
DATABASE_URL	(derived)	Full async DB URL (overrides individual vars)

Example: Store and search vectors

# Store a vector
curl -X POST http://localhost:8000/vectors/ \
  -H "Content-Type: application/json" \
  -d '{"label": "example", "content": "some text", "embedding": [0.1, 0.2, ...]}'

# Search for similar vectors
curl -X POST http://localhost:8000/vectors/search \
  -H "Content-Type: application/json" \
  -d '{"vector": [0.1, 0.2, ...], "top_k": 3}'

Note: The default embedding dimension is 1536 (OpenAI text-embedding-ada-002). To change it, edit VECTOR_DIM in app/models.py before the first run.