Linking a place authority record to a usable geographic shape is often a manual, time-consuming step in collection management systems — yet the linked data is frequently already there. This lightning talk demos a small, open, dependency-free (Python standard library only) pipeline built for Axiell Collections at the Rijksmuseum, which takes a Wikidata QID, follows its OpenStreetMap relation (P402) or point coordinate (P625), fetches the geometry via Overpass, and converts it into a ready-to-import Axiell record — with no manual GIS work required. The real value is in the edge cases: real OpenStreetMap and Wikidata data rarely behaves as cleanly as the happy path suggests, and this talk walks through a few concrete examples of how the pipeline handles them without silently producing a wrong result. It also processes hundreds of records in a single batch, with a clear audit trail of what needs human review. These are exactly the kinds of real-world data quirks any institution connecting Wikidata/OSM to a legacy collection system will run into. Attendees will leave with a concrete, reusable pattern for this kind of linked-data-to-CMS pipeline, regardless of which collection management system they use.
