Agentic GIS for Google Cloud with CARTO
The first geospatial agent in the Gemini Enterprise Agent Gallery
CARTO is the first and only GIS company to have passed Google's highest governance, security, and accuracy standards for its AI agent. Curious to learn more? Visit the Agent Gallery inside the Gemini Enterprise app and try one of our pre-built agents, ready to use in seconds.
“To deliver real impact from agentic AI, customers need a simple, trusted way to interact with specialized domain data using intelligent agents like the CARTO Site Selection AI Agent. By making this agent accessible through Gemini Enterprise, we're giving customers greater flexibility to apply spatial intelligence to their expansion strategies and turn complex geospatial data into meaningful business value.”
Two ways to adopt Agentic GIS with Gemini Enterprise
Some teams live in CARTO. Others live in Gemini Enterprise and the AI tools they already use. Agentic GIS covers both, and the analysis always runs on your governed data in BigQuery.
Build inside CARTO with Google AI
Get more from your geospatial analysis with Google's AI capabilities, directly in the CARTO platform. Use natural language, automate workflows, and build Location Intelligence applications faster.
- CARTO AI Agents answer plain-language questions on Builder maps
- AI Assistants speed up spatial SQL, maps, and Workflows
- Connect to the Gemini Enterprise Agent Platform and keep your guardrails
Build outside CARTO on your favorite AI
CARTO for Agents brings spatial analysis into the agent platforms your teams already use, such as Gemini Enterprise. Your answers now include maps, grounded in workflows built by your GIS team, without opening CARTO.
- Try a pre-built agent on the Gemini Enterprise Agent Gallery
- Connect through the CARTO MCP Server and the CARTO CLI
- CARTO Agent Skills ground results in geospatial know-how
Agentic GIS in action with Google Cloud
.webp)
Frequently asked questions
What is agentic GIS, and how does it work on Google Cloud?
Agentic GIS is spatial analysis run by AI Agents instead of by hand. On Google Cloud, CARTO agents connect to your data in BigQuery, run the analysis, build maps, and automate Workflows, all inside the security and governance you already have. Analysts move from running every query themselves to directing agents that do the work.
How do AI agents run geospatial analysis in BigQuery?
CARTO AI Agents work directly on your data in BigQuery. They read your tables, run spatial functions and Workflows, and return maps and answers without moving data out of the warehouse. Because they call defined Workflows rather than generating code on the fly, results are repeatable and auditable.
How does CARTO work with Gemini Enterprise for geospatial analysis?
CARTO built the first geospatial agent in the Agent Gallery inside the Gemini Enterprise app, validated against Google's security and accuracy standards. Teams ask location questions in Gemini Enterprise and get answers backed by real spatial analysis running in BigQuery.
Can I use CARTO from my own AI tools, without opening CARTO?
Yes. CARTO for Agents brings spatial analysis into the agent platforms your teams already use, through the CARTO MCP Server, the CARTO CLI, and CARTO Agent Skills. Ask a question in Gemini Enterprise or another agent platform and the analysis runs on your data in BigQuery, returning maps and answers grounded in workflows built by your GIS team.
Which LLMs can I use for geospatial AI agents on Google Cloud?
You bring your own models. CARTO works with the Gemini Enterprise Agent Platform and Google AI Studio, so your models and your data stay in Google Cloud. There is no vendor lock-in, and the same agents work with other providers through open standards.
Can I use Google's geospatial foundation models in CARTO?
Yes. As a Gemini customer, you can access Google's geospatial foundation models directly through CARTO, including Earth Engine satellite embeddings and Population Dynamics (PDFM) embeddings, and use them in your existing geospatial workflows. You get insight from satellite imagery and other spatial data without building and training machine learning models from scratch.
How are AI agents kept secure and governed in BigQuery?
Agents run inside BigQuery and inherit the same permissions and roles as your team, with scoped, short-lived access instead of standing service accounts. Because they call defined Workflows rather than generating SQL on the fly, every action is repeatable and logged, so you can audit what an agent did and why.