Cloud Pak for Data

Cloud Pak for Data

Come for answers. Stay for best practices. All we’re missing is you.

 View Only

Integrating Claude with Watsonx Data Intelligence MCP Server

By RAMAKANTA SAMAL posted 12 days ago

  

Introduction

Enterprise AI workflows increasingly rely on agent frameworks and tool orchestration to connect foundation models with business data. IBM’s Watsonx Data Intelligence offers metadata-driven governance and compliance capabilities, while Claude brings advanced reasoning and conversation skills.

By integrating these with the Watsonx Data Intelligence MCP Server, you can allow Claude to directly access governance tools such as rule creation, search, and lineage queries — all in a secure, structured way.

Architecture Overview

Here’s the high-level flow:

  1. Claude MCP Client (Claude Desktop)– Runs locally,  invoking MCP servers.

  2. Watsonx Data Intelligence MCP Server – Wraps Watsonx Data Intelligence APIs as structured tools.

  3. Watsonx Data Intelligence Services – Back-end services such as rule service, search service, lineage service.

Prerequisites

Before starting, make sure you have the following installed or configured:

Steps: 

Configure IBM watsonx data intelligence mcp server with claude desktop.

  1. Open claude desktop app
  2. Go to settings -> Click on Developer -> Edit Config
Open the file claude_desktop_config.json in a text editor and and add a new entry under mcpServers as shown below.

Please replace the corresponding api key, data intelligence service url and username. For SaaS env we don't require the username.

Below is the configuration for SaaS Data Intelligence:

{
  "mcpServers": {
    "wxdi-mcp-server": {
      "command": "uvx",
      "args": [
        "ibm-watsonx-data-intelligence-mcp-server",
        "--transport",
        "stdio"
      ],
      "env": {
        "DI_SERVICE_URL": "https://api.dataplatform.cloud.ibm.com",
        "DI_APIKEY": "<IBM IAM api key>",
        "DI_ENV_MODE": "SaaS"
      }
    }
  }
}
Below is the configuration for CPD onprem Data Intelligence:
{
  "mcpServers": {
    "wxdi-mcp-server": {
      "command": "uvx",
      "args": ["ibm-watsonx-data-intelligence-mcp-server", "--transport", "stdio"],
      "env": {
         "DI_SERVICE_URL": "<cpd DI service url>,
         "DI_APIKEY": "<cpd  api key>",
         "DI_ENV_MODE": "cpd",
         "DI_USERNAME" : "<cpd user name>
      }
    }
  }
}
Close the Claude Desktop app and reopen it. You should now see the MCP server listed under your Claude Desktop app. And make sure its enabled.
Now you can start asking questions related to data intelligence services:

Conclusion

By integrating Claude with Watsonx Data Intelligence via MCP Server, enterprises can enable AI-driven governance workflows without compromising control. This setup allows Claude to go beyond general reasoning and act as a secure assistant for governance tasks like rule creation, asset search, and lineage queries.

0 comments
31 views

Permalink