Demonstrate how to tie a database into the upstream data pipeline

1. Add a build target that populates a database with some random data
2. Add a build target that writes to a file to match the most current last_updated_date for any database tables we're interested in.
3. Add this file as a data dependency

... or something like that. Could trivially add the db file as a dependency, but let's pretend we only have network access to the database to prove the concept.