Configuration of Hasura is usually done via the HasuraConsole. However during development - especially in case of a Spring web application using Hibernate as the ORM - the database schema may change a lot and it is tedious to always set table and relationship tracking, permissions etc. on a newly generated database in the HasuraConsole. With HasuraConfigurator when the data model changes in Java the HasuraConfigurations can be regenerated and automatically loaded into Hasura thus the data model and the HasuraConfiguration will always be in sync.
As an added bonus, HasuraConfigurator also makes it very easy to add various customisations on the model, like defining aliases for column names to match their Java equivalent, or renaming root operations so that they look better (ie. camelCase instead of snake_case) in the graphql schema.