AI assistants and agents are only as useful as the data they can reach. Ask one about last quarter’s orders, and it can only answer if it can see the system where those orders live. For most enterprises, that system isn’t a modern cloud database. It’s a mainframe, a VSAM file, or a legacy database that has quietly run the business for decades, and that AI tools can’t see.
That’s the gap the new CONNX MCP Server Community Edition is built to close. It gives AI tools secure access to the data sources you already connect through CONNX, from mainframe files to relational databases, without moving or copying that data first. Your AI assistant gets a direct, governed path to live enterprise data, and your teams get answers grounded in the systems of record they already trust.
What is MCP?
Model Context Protocol (MCP) is an open standard that lets AI applications like Claude, ChatGPT, and Copilot connect to external tools and data in a consistent way. Think of it as a common plug: instead of building a custom integration for every AI tool, a data source exposes one MCP server, and any MCP-compatible client can use it. For a deeper look, see our MCP explainer.
What the CONNX MCP Server does
The CONNX MCP Server sits between your AI tools and the trusted data access layer that CONNX already provides. In practice, that means three things:
- It exposes your CONNX-connected data sources to MCP-compatible AI clients. Once it’s running, tools like Claude, ChatGPT, and Copilot can discover and query those sources through a single, standard interface.
- It lets users ask questions in plain language and get answers from live data. No SQL, no extract jobs, no waiting on a report. The AI translates the question, CONNX retrieves the data, and the answer reflects what’s in the system right now.
- It works with the sources you already connect through CONNX, legacy and modern. VSAM, IMS, Adabas, and other mainframe sources sit alongside relational and cloud databases, with no new connectors to build and no data to migrate.
Why it matters: From question to answer in minutes
Picture a financial analyst preparing for a quarterly review. She needs last quarter’s orders, broken down by region. Those orders live in VSAM files on the mainframe.
Today, getting that data usually means filing a request, waiting for someone with mainframe skills to write an extract, and then reconciling a spreadsheet that’s already a few days old. With the CONNX MCP Server, she opens her AI assistant and asks: “Show me last quarter’s orders by region.” The assistant queries VSAM through CONNX and returns the numbers in seconds.
That simple exchange captures three things that matter to the business:
- Speed. Questions that used to take days of back-and-forth get answered in the moment, while the decision is still on the table.
- Trusted data. The answer comes straight from the system of record, not from a copied dataset that may be stale or incomplete. Access runs through CONNX’s existing security, so users see only the data they’re permitted to see.
- AI that scales across the enterprise. Because MCP is a standard, the same server works across AI clients. Once your data is connected, every team can put it to work, with no custom integration for each new tool.
Comece agora
The scenario above is just one of many that the CONNX MCP Server can support. Your most valuable data shouldn’t be off-limits to your AI tools just because it lives on a legacy system. Sign up for the CONNX MCP Server Community Edition and connect your first AI assistant to the data that runs your business.