CARTO AI Agents now integrated with Amazon Bedrock
.png)
Spatial analysis has long played a critical role in answering the “where” and “why” questions that shape strategic business decisions — from planning telco networks and optimizing logistics to understanding human mobility and environmental risk.
Yet these insights have often been siloed, requiring specialized tools and expertise that limit their accessibility across an organization
With the launch of Agentic GIS, CARTO introduced a new era of Location Intelligence where geospatial analysis is interactive, intuitive, and accessible.
Today, we are thrilled to extend that vision and announce CARTO’s latest Agentic GIS integration with Amazon Bedrock. This allows users to power geospatial AI agents and assistants using Large Language Models (LLMs) available through AWS services, fully aligning with existing cloud infrastructure, security, and governance.
Agentic GIS is CARTO’s AI-native approach to geospatial analytics, designed to move beyond the limits of traditional GIS systems. Instead of being constrained by complex software or siloed tools, your analytical workflows become interactive and accessible across your entire organization.
Our vision for Agentic GIS is built on three pillars:
- Reach: AI Agents make complex geospatial insights accessible to everyone by scaling geospatial team expertise into automated workflows. These agents are capable of planning and executing multi-step geospatial workflows— like catchment analysis or spatial joins— based on high-level goals provided by business stakeholders in natural language.
- Velocity: Embedded AI Assistants boost the productivity of the GIS analysts by accelerating everyday tasks, from generating spatial SQL and summarizing datasets to building maps and workflows.
- Depth: Thanks to Geospatial Foundation Models trained on vast amounts of spatial data, enabling a deeper understanding and discoverability of spatial patterns and insights.
Now with our new Amazon Bedrock integration, you can power CARTO’s Agentic GIS using LLMs you already trust — governed and managed directly in AWS Bedrock..
But it goes a step further.
Through our MCP Server, you can expose your CARTO workflows as accessible MCP tools. This enables your AI agents to run and operate fully on AWS, scaling your team’s geospatial expertise across the organization, without ever leaving your secure AWS cloud environment.
For joint AWS and CARTO users, this native integration unlocks a powerful, unified workflow. It’s designed to help you:
- AI-powered geospatial analytics: Run AI-powered geospatial analytics using the foundation models available through Amazon Bedrock you already use.
- Maintain unified governance and security: Your data never leaves AWS. This integration is built to respect all your existing data access policies and security controls implemented in AWS.
- Leverage cloud-native spatial processing on AWS: All spatial processing and computation in CARTO is executed directly where the data resides — Amazon Redshift, Databricks, Snowflake, Oracle or PostgreSQL-based databases — ensuring optimal performance, scalability, and consistency.
- Build composable agentic experiences: Expose your spatial workflows as MCP Tools, allowing AI Agents in CARTO to seamlessly connect to your Bedrock-managed LLMs in your existing environment.
As an AWS Technology Partner, CARTO is built to run natively on AWS, making it easier than ever to embed trusted GIS processes into modern analytics and AI initiatives.
Setting up CARTO AI with AWS Bedrock is designed to be straightforward.
Select Amazon Bedrock on the CARTO AI page in Settings. Then enter the following configuration details to access AI models directly in your AWS environment:
- AWS Access Key ID
- AWS Secret Access Key
- AWS Region
To see models supported check out our documentation here!
Don’t have access to CARTO Settings? Contact your CARTO Admin to enable.
If you’re using CARTO on AWS, you can now enable Agentic GIS with Amazon Bedrock and start building geospatial AI Agents with the foundation models you trust.
Ready to try geospatial AI in AWS? Request a demo with one of our experts or explore our documentation to start with Agentic GIS powered by Bedrock.
.png)
