Raton is from-scratch rebuild of my Yenot framework. The goal is to tie parameterized HTTP routes to Python functions with built-in connection to a PostgreSQL database. A library of SQL convenience functions for read and write are included. These support annotating results sets with PG types and writing deserialized HTTP bodies to the database.
The goal is not to be an ORM but to provide the tools that allow and encourage direct access to raw SQL.
Raton is built on a few carefully selected foundational libraries including primarily Starlette and Psycopg.
There is nothing fundamentally linked to Docker in the architecture of Raton but containers have proven to be a convenient development tool and the deployment documentation assumes a docker environment. The included docker-compose.yml may be a useful template to pattern for your own deployment but it is recommended build your own. For one certainty you will need to provide some distinction in your environmental configuration for the database access.