// the find
vargasjona/fastapi-alembic-sqlmodel-async
This is a project template which uses FastAPI, Pydantic 2.0, Alembic and async SQLModel as ORM. It shows a complete async CRUD using authentication and role base access control.
A FastAPI project template wiring together SQLModel, Alembic, async SQLAlchemy 2.0, JWT auth, RBAC, Celery, Redis, and Minio — all running behind Caddy in Docker Compose. It's aimed at developers who want a pre-assembled Python backend stack rather than building one from scratch.
UUID7 primary keys are a thoughtful default — time-sortable and Postgres-native without needing a separate library at runtime. The CRUD base class pattern with SQLModel + async sessions is clean and avoids the boilerplate most projects paste everywhere. Celery Beat tasks stored in Postgres via celery-sqlalchemy-scheduler is a practical choice that avoids Redis as a second source of truth for schedules. The pg_trgm-based smart user search is a nice touch that most templates skip.
Testing is listed as TODO — there are skeleton test files but no actual coverage, which means the template ships without the thing that would let you trust it when you start changing it. SQLModel still leaks SQLAlchemy in ways that bite you on complex queries; the template doesn't address this and a new adopter will hit it fast. The Python version pin is '>3.9,<3.12' in the README but the code uses 3.10+ syntax — that range is wrong and will cause confusion. The Langchain+OpenAI chatbot example is bolted on and uses a different AI vendor than the rest of the stack, which makes it feel like a feature grab rather than a coherent architectural choice.