Skip to content
Get Started

FME with MCP: Extend the Reach of AI Across Your Enterprise.

FME with Model Context Protocol (MCP) is a governed connection layer between any AI model, tool, or application and the hard-to-reach data most integrations can’t touch. Securely connect your data, without lock-in or code, and gain complete enterprise context with control over the workflows, data, and actions AI can access.
Webinar
Inside the Firewall: MCP-Powered Paths to Enterprise AI That Respect Your Data
Watch On Demand
In this webinar you will learn how to:
Connect leading cloud AI agents to your on-premises spatial and operational data through MCP.
Filter what’s exposed so AI gets the answers it needs while sensitive information stays behind the firewall.
Pair on-premises models like Gemma 4 with MCP Servers hosted on your internal FME data infrastructure.

Connect your AI, tools, and applications to thousands of systems.

FME with MCP provides a secure, live connection to spatial, legacy, on-premises, and other hard-to-reach data sources for complete enterprise context.
Learn More
MCPserver-fmeflow

FME with MCP brings choice and control to enterprise AI.

Give AI Complete Enterprise Context

Connect spatial, legacy, on-premises, real-time, unstructured, and other hard-to-reach data that other integrations can’t touch without code.

FME enables a live connection to your source data so every AI model you connect draws on complete, current context instead of a stale snapshot or a partial view.

Choose What AI Can See and Do

FME workspaces define exactly what data is exposed to AI, filtering data at the source so AI only ever sees what you choose to expose.

Trusted by Design

Jobs run on FME Flow with role-based permissions, a full audit trail, and OAuth 2.0-protected access, giving IT and security teams the same governance over AI access that they already have over every other system connected to FME.

Deploy Where Your Data Lives

Deploy FME wherever your data already lives, whether in the cloud cloud, hybrid, on-premises, or fully air-gapped. Run the model, client, and server entirely on-premises so sensitive data never crosses onto the public internet, and MCP access to it is enforced through the same infrastructure your compliance team already trusts.

Extend AI’s Reach with FME’s Capabilities

Any MCP-compatible client can call FME’s readers, writers, transformers, and native spatial analysis to transform data, orchestrate multi-step workflows, or take action in systems AI couldn’t otherwise reach.

Choose Today, and Again Tomorrow

Pick any MCP-compatible AI and switch as technology evolves, without rebuilding your integrations.
FME MCP Diagram DRAFT6 Use MCP Tools horizontal BLUE (2)

MCPCaller Transformer

The MCPCaller Transformer is the bridge between FME and MCP tools. Its dual-mode architecture supports deterministic workflows, where execution is predefined, and dynamic execution, where the best action is determined at run time.

Design-Time (Deterministic)

The FME author explicitly specifies the MCP tool to use, ensuring precise execution for workflows in which the tool to be executed is known in advance.

Run-Time (Dynamic)

The FME workspace evaluates intent at run time and selects the most appropriate tool from the available MCP inventory, enabling autonomous, context-aware execution.

Unified Schema

Both modes share the same underlying feature set, allowing tools configured at design time to be switched to dynamic execution easily.

Adaptive Error Handling

FME manages the inherent unpredictability of external tool environments. When a tool fails for any reason, a controlled fallback allows authors to define “safety-net” pathways.
FME MCP Diagram DRAFT6 Create MCP Tools horizontal BLUE (3)

FME Flow’s MCP Server

FME Flow’s MCP Server exposes FME workspaces as MCP-compatible tools that any AI model, application, or tool can securely discover and invoke, protected by OAuth 2.0 and governed by the same role-based permissions and audit trail as any other job on FME Flow.

It gives any MCP-capable client access to FME’s full capabilities, including its readers, writers, transformers, and native spatial processing.

Publish Existing Workspaces as AI Tools

Expose your existing FME workspaces as MCP tools that AI, applications, and agents can securely discover and invoke, without creating custom APIs or integrations.

Value from Day One

Use curated, out-of-the-box tools to start solving problems immediately.

Maintain Existing Security and Governance

Integrate with OAuth 2.0 and your FME Flow server for authentication, permissions, validation, logging, and execution. Deploy your model, MCP client, and FME Flow entirely on-premises so governed data remains inside controlled infrastructure.

Become an MCP Tool Provider Without Code

Simply publish an FME workspace to create an MCP tool, with no code and no new skills required.

Discover how FME with MCP could work for you.

Explore use cases to see how FME with MCP connects AI to enterprise data and workflows to drive measurable outcomes.
Nottingham Council House at night
Citizen Services & Planning Support

Public Sector

Challenge: Government agencies face overwhelming citizen inquiries and growing backlogs for permits, zoning, and service requests. 

Solution: FME Flow’s MCP Server securely exposes approved government systems and policy data through governed, access controlled endpoints.

Benefits: Agencies can provide 24/7 support without exposing sensitive data, ensure requests receive accurate information, and reduce the manual workload placed on GIS and IT teams.

MediumJPG workers night laptop powerlines
Outage Response & Asset Operations Support

Utilities & Energy

Challenge: Utilities must respond quickly to outages and asset failures, but critical information is spread across GIS, asset management, SCADA, and maintenance systems, making it difficult to get a complete, reliable picture when time is critical. 

Solution: FME workflows use the MCPCaller to invoke approved MCP tools that retrieve and validate asset registry, work order, IoT, and compliance data in real time.

Benefits: Teams can reduce manual data lookups across systems and answer operational questions quickly using trusted, up-to-date information.

Lightrail trains
Construction & Traffic Operations Support

Transportation

Challenge: Transportation agencies coordinate construction activities and traffic impacts, but fragmented data across systems makes it difficult to answer operational questions quickly.

Solution: FME Flow’s MCP Server exposes new OAuth 2.0-protected MCP tools that retrieve construction schedules, query asset inventories, and assess traffic impacts across systems, providing a consistent and authoritative operational view.

Benefits: Teams can understand construction and incident impacts faster, reduce manual data compilation, and provide accurate information to internal teams and the public.

Frequently Asked Questions

FME with MCP makes it easy to give AI agents secure, controlled access to your systems and data. Instead of building custom integrations or relying on stale training data, models can interact directly with your APIs, services, and workflows, accelerating time-to-value while staying governed and predictable.

By running an MCP server, you explicitly define what an AI can do in your environment—whether it’s to retrieve data, trigger actions, or run searches. This extends the model’s capabilities safely without any guesswork or hallucinations, because the AI only operates within the boundaries you’ve set.

Traditional function calling and tool integrations are typically built directly into an application and tied to a specific AI model or provider. These work well for single use cases, but they don’t scale as easily across multiple AI models, applications, or environments.

FME with MCP standardizes this integration layer by securely exposing workflows, data, and tools in a way any AI model can understand. This allows organizations to connect any AI to any system without vendor lock-in and without code, and to reuse those capabilities across models as technology evolves.

MCP capabilities are now available in FME. The MCPCaller Transformer is available in FME 2026.1+, and FME Flow as an MCP Server is available in FME 2026.2.

FME with MCP will be included at no additional cost for customers with FME Form and FME Flow.

Additional Resources

MCP

Webinars

Resources

Questions about the platform?

We’re here to help you find the answers and information you need to continue your data integration journey.
Contact Us

Pricing, with a difference.

Valuing long-term partnerships over short-term profit, we tailor our pricing to each individual client.
Explore Pricing

Master FME with our interactive demos.

Learn the platform and unlock FME’s full potential. Bring life to your data, today.
Discover FME Demos