Natural language meets point cloud: Building an AI agent for spatial data

Spatial data is becoming conversational – and the GIS teams that understand this shift will have a significant advantage.

In this session, I’ll demonstrate a working AI agent that queries and even modifies real point cloud and spatial datasets through natural language, handling the kind of complex, multi-step reasoning that off-the-shelf AI tools can’t manage.

Beyond the demo, you’ll walk away with a clear picture of what this architecture means for your own product or workflow, and how to approach it practically – no hype, no theory, just a working system and honest lessons from building it.