The open-source geospatial community has built powerful tools that manage and analyze massive datasets with unprecedented speed and steadily decreasing cost. These tools now support analyses at continental and global scales and increasingly contribute to the training of AI models.

Despite the focus on scale, we often overlook the foundational geographic data infrastructure that provides essential analytical context. This oversight can introduce ambiguity and hinder the interoperability of our data and services. For instance, what do we actually mean by a “state” or a “dam”?

This talk highlights the essential role of data architecture, linked data, ontologies, and metadata in geospatial analysis. It will show how applying FAIR principles - making data findable, accessible, interoperable, and reusable - can improve quality, enhance human understanding, and enable machine-readiness. These elements form the foundation of effective, AI-driven analysis in the open-source geospatial ecosystem.