Supplier data came from several places.
It arrived from other systems and from manual entry. The goal was to let leaders see in one place which suppliers confirmed forecasts on time, and where production stalled because of it.
The reports were built on a single data model.
AWS Lambda functions processed the data and loaded it into a star schema in PostgreSQL, which the reports were built on. The application itself was a monolith following Clean Architecture, with a React front end, released through GitHub Actions to Kubernetes.
Leaders could follow their suppliers in one place.
They could filter the figures by business unit, region and supplier, compare successive forecasts part by part, and get a daily email summary of the suppliers they followed.