When starting a new project that uses a database, I simply add a new Express Router and define the endpoints. Endpoint handlers normally handle the logic for sanitizing / typechecking inputs before piping the inputs (if any) through function calls on a Database Utility class. This class is responsible for making the actual API calls on the database framework being used, in this case, SQLite3. The result of this operation is then returned back to the requesting party as an HTTP response.

     An instance of this application runs on a server I host, and it is the component responsible for maintaining Tune Mountain’s backend needs, as well as WPI Map Poll’s responses’ storage location. It also holds data for other Data Visualization projects.

     I like to note that any optimizations that can be done for minimizing large data bandwidth use are done here. As an example, calculations for averages / medians / etc for the WPI Map survey responses are done in the back-end so thousands of bytes of unnecessary repeated data aren’t sent to the end user.