Every summer, the Esri User Conference brings together tens of thousands of GIS professionals to share what's working, what's next, and where the technology is heading. This year's UC in San Diego did not disappoint.
Three themes dominated the conversation, and they're deeply connected.
Agentic AI is no longer theoretical.
AI tools that don't just analyze data but act on it were front and center at this year's UC. Esri's platform is increasingly designed to support AI agents that can query, interpret, and respond to spatial data without constant human prompts. For government GIS programs, the conversation is shifting from "can we use AI?" to "is our data ready for AI?"
Model Context Protocol (MCP) changes how AI connects to data.
MCP is an emerging standard that lets AI models tap directly into live data sources, including GIS systems. Think of it as a universal connector between AI tools and your operational data. The organizations that have invested in well-maintained, structured GIS data will be positioned to take advantage of these integrations far more quickly than those still working off legacy formats.
Digital twins require a foundation.
Digital twin projects were everywhere at UC. But the message from practitioners was consistent: the technology is ready, the data often isn't. Accurate parcel layers, current utility as-builts, clean address points, the fundamentals matter more than ever when you're trying to build a living model of your community.
The through-line across all three themes: data readiness is no longer a nice-to-have. It's the prerequisite.
Strategic planning is what makes adoption possible.
The pattern we see consistently: the clients who are actually taking advantage of technologies like digital twins didn't get there by accident. They started with a GIS Strategic Plan, aligning stakeholders, establishing data governance, and building a roadmap tied to real organizational goals. That foundation is what makes the difference between watching a trend and being able to act on it. When the tools arrive, the groundwork is already done. The clients rushing to stand up digital twins today are largely the ones who made that planning investment years ago.
Connected data is what makes AI useful.
One of the most overlooked barriers to AI in government GIS is straightforward: the data is trapped. Operational data sitting in an ERP or permitting system can't be queried by an AI tool, can't be visualized on a map, and can't power the spatial analysis that makes GIS valuable in the first place. GeoAdaptor™, GTG's new middleware product, is built specifically to solve this. It connects non-spatial operational databases directly into your GIS environment, so the data your agency already has can actually be put to work. AI is only as good as the data it can reach. Getting your systems connected is the first step.
GTG's team came back from San Diego energized. The work we do across the full GIS lifecycle from strategic planning through implementation and ongoing support — is exactly the kind of foundation these new capabilities are built on. If you've invested in getting your GIS right, the tools coming out of UC are going to work for you.