Safely Connect Agents to the Work That Runs Your Business
InterSystems IRIS® AI Hub™ provides a comprehensive set of capabilities for building, deploying, and governing AI agents and agentic workflows. Embedded within InterSystems IRISand InterSystems IRIS for Health®, it provides a unified environment for developing agentic AI applications that are tightly integrated with your enterprise data, business logic, and operational workflows. Teams can build native AI agents, safely connect external agents to governed tools and live operational context, publish business logic as governed tools using the Model Context Protocol (MCP), and apply consistent security, governance, and human oversight throughout the AI lifecycle, whether your data already sits in InterSystems IRIS or across your solution ecosystem.
Closing the Gap Between an AI Pilot and Production
Calling a language model is easy. Putting one into production is not. The moment an agent needs live enterprise data, permission to act, and an accountable record of what it did, most teams discover their AI stack sits outside the controls their business runs on. Credentials are spread across cloud runtimes. Data is copied to wherever the AI lives. Each new agent becomes a fresh security review and risk.
AI Hub closes that gap by making the data platform the place where AI, built anywhere, connects safely to the business. Agents work from live operational context, use only the tools they are approved to use, and operate under the same security, workflow, and audit controls your applications already run on.

Five Capabilities. One Platform. Your Governed 360° View.
Build Native Agents Alongside Your Applications
AI Hub offers familiar patterns to build agents and agentic workflows quickly on InterSystems IRIS. With convenient abstractions for language models, prompts, tools, and skills, you can create native agents quickly, with direct, governed access to existing data assets and business logic.
The resulting agents fit perfectly in your existing software development lifecycle and can ship and run with your mission-critical applications without requiring additional installations, separate configurations, or DevOps workflows.
Ground Agents in Live Operational Context
Direct access to the most relevant and current information is critical for agentic applications. Tightly integrated with InterSystems IRIS and running right where your data lives, AI Hub gives agents the complete picture of live records, events, transactions, and process state.
Connect Agents to Business Tools with Bidirectional MCP
To make agents act rather than just answer, they need access to tools that are relevant to the current business. Agents can access native tools directly, and exposing business logic as MCP tools takes just a few clicks. Agents call approved capabilities with explicit contracts rather than receiving raw database or API access. MCP works in both directions: external agents and AI clients can use the IRIS tools you approve, and native agents can call external business tools.
Govern Agents with the Controls You Already Run
AI Hub lets you reuse your existing roles, access policies, and credential management, so that you can govern your agentic and human users in a common language, with every action auditable. These controls apply to the resources IRIS manages; connected applications continue to enforce their own.
Orchestrate Agents, Deterministic Processes, and People
Agents interpret and propose, established IRIS processes execute under deterministic rules, and people approve wherever policy requires, so agents can take part safely in long-running, mission-critical processes.
For decades, organizations in regulated industries have trusted InterSystems to power mission-critical, long-running transactional- analytic connected applications. That experience is why InterSystems is well-positioned to help you safely connect agents to the work that runs your business. AI Hub is built into IRIS, so agents work from live data, use only approved tools, and take part in your existing workflows through native MCP support and IRIS interoperability. Keep the access controls your applications already use, and continue to leverage your existing models, agent frameworks, and enterprise applications.









































