GraphQL simplifies life for clients because there is a single HTTP endpoint. In some ways, this makes the pattern for a serverless GraphQL API extremely simple and in some ways quite dull.
If we were starting this GraphQL web application from scratch, there would be plenty of decisions to make and material to cover to make our application code modular, easy to test, and well designed. Since we're porting the example REST web application, we have already implemented the vast majority of the needed functionality and software layers. These sections may seem terser than expected, especially if you have skipped Chapter 2, A Three-Tier Web Application using REST. Any gaps in code organization or layout, configuration strategy, deployments, ...
