How to use graphene dataloader on connections and how to select only necessary fields

Question

I would like to use some non-standard database with python graphene to create a GraphQL endpoint. This database doesn\'t have existing SQLAlchemy dialect, is not supported by Dj