AI Agents

This page covers in-product AI Agents — agents you build inside CARTO Builder maps and publish to your organization. Looking to drive CARTO from an external agent like Claude Code, Claude.ai, ChatGPT, Cursor, or Gemini CLI? See CARTO for Agents.

CARTO AI Agents provide a powerful conversational interface that allows anyone, regardless of technical expertise, to ask questions in natural language and receive instant, actionable insights. This marks a fundamental shift beyond dashboards to a dynamic, intuitive way of exploring your geospatial data.

With CARTO, you can create and publish powerful geospatial AI Agents tailored to your specific needs. By combining your custom prompt instructions with CARTO's built-in tools and your own MCP tools, you can build trustworthy solutions that make complex geospatial analysis accessible to any user within your organization.

What is an AI Agent?

An AI Agent is a sophisticated system powered by a large language model (LLM) that goes beyond simple chat. It autonomously interacts with your data and a robust set of tools to solve complex geospatial tasks, reasoning through problems to deliver actionable, data-driven insights.

Each agent you create is built from three core components:

• Logic (Use Case & Instructions): The agent's "brain" — defining its purpose, behavior, and expertise through clear, structured instructions.

• Tools & Capabilities: The agent's "hands" — accessing a powerful suite of built-in geospatial tools and extending skills with your custom MCP tools.

• Model (LLM): The agent's "engine" — powering reasoning, language understanding, and decision-making capabilities.

Getting started with AI Agents

This documentation walks you through the complete journey of building and deploying AI Agents:

Learn the basics

• Creating your Agent - Set up your agent and understand the interface

• Agent Config Assistant - Configure your Agent through conversation: create, iterate, and refine

• Understanding Agent behavior - Learn how agents interpret queries and make decisions

• Choosing the right model - Select the optimal LLM for your use case

Build and configure

• Defining your Agent logic - Write effective Use Cases and Instructions

• Working with tools - Integrate Core Tools and custom MCP Tools

• Configuring capabilities - Enable advanced features like Query Sources

Deploy and share

• Iterate and refine - Test and improve your agent based on feedback

• Sharing your Agent - Publish and distribute to your organization

Reference

• Tools Reference - Comprehensive documentation of all available Core Tools

Organization Admins can monitor AI Agent activity (conversations, MCP tool executions, and top users and agents) from the CARTO AI Analytics tab in Settings.