@jure That is one of the goals for graphql, hide to the frontend the backend complexity behind the graphql queries. You can create your schema with different datasources. In that case the complexity must be managed by your resolvers.