sqlalchemy-model-generator

What problem does it solve?

Create structured SQLAlchemy models (with UUID primary keys) and optional vector embeddings, enabling PostgreSQL-based vector search and BM25 indexing for fast retrieval in ML-assisted apps.

Core Features & Use Cases

• Auto-generate model classes following existing conventions (Base, tablename, timestamps, UUID PK).
• Add embedding columns using pgvector (Vector(dim)) with dimension controlled by settings.
• Implement full-text search support via TSVector and appropriate indexes for BM25 lookups.
• Update initialization and retrieval scaffolds (init_db.py, retrieval service) to integrate new models seamlessly.
• Read and reference existing models to maintain consistency and lineage.

Quick Start

Provide the table name and field specifications to generate a new SQLAlchemy model with pgvector embedding and BM25 indexing.