This JUG recording of a live coding presentation shows how to build an AI planning optimization application from scratch on Quarkus that optimizes complex timetabling problems with OptaPlanner.
It exposes a REST service and connects to a database. By the end, constraints are added one by one - causing the "Solve" button to generate radically different schedules after each addition — improving the lives of users as more constraints are taken into account.
The source code build for this presentation is available in GitHub.