Local governments connect AI chatbots to GIS for location-based answers

GIS tools are becoming part of government chatbot workflows rather than remaining separate map interfaces. Maricopa County’s agents combine text retrieval with spatial operations, including address-to-point conversion and nearest-location searches.

Residents can find nearby resources without learning a GIS interface, while government teams must ensure conversational answers accurately reflect official datasets. Carrying addresses and parcel context across questions also makes privacy controls and conversation retention operational concerns.

Local government chatbots are beginning to connect conversational interfaces with spatial tools, allowing residents to ask where services and resources are located rather than navigate maps or datasets themselves. In Maricopa County, public-facing Mari and employee-facing Marco combine IBM Watson conversational models, retrieval-augmented search and Esri GIS tools. The county formally launched the system in June 2026 after testing 11,000 prompts. Mari can turn an address into a map point, calculate nearby resources and carry location context into follow-up questions. Washington, D.C., has also deployed DC Compass for its open data portal, following a pilot launched with Esri in March 2024. These deployments make GIS functionality accessible through ordinary questions, but put greater weight on authoritative data and controls around personal information. The county reports access restrictions, retention limits and guardrails; the supplied material does not establish independent measures of answer accuracy.

Evidence supplied by Keely Quinlan, StateScoop.