CData Connect AI (Preview)

Connect your agents to read, write, and act on 350+ enterprise data sources like Salesforce, Snowflake and NetSuite in real-time through one fully hosted remote MCP server. Connect AI is the only managed MCP platform built with semantic intelligence that understands your data's metadata and relationships across systems for multi-source analysis. Ensure trusted boundaries with enterprise security, granular permissions, and complete audit trails. Set up any source in minutes and get started free.

This connector is available in the following products and regions:

CData Connect AI MCP Server

Overview

The CData Connect AI MCP Server enables AI agents and copilots to interact with your enterprise data sources through the Model Context Protocol (MCP). This connector provides a secure gateway to query and manipulate data across multiple connected data sources including databases, SaaS applications, and cloud services.

Publisher

CData Software Inc

Prerequisites

Before you can use this connector:

1. Connect a data source to your CData Connect AI account. See Sources documentation for more information.

Authentication

This connector uses OAuth Authentication with the following format:

• Username: Your Connect AI email address
• Password: Your Connect AI password

Available Tools

The CData Connect AI MCP Server provides the following tools for data discovery and manipulation:

Data Operations

• queryData: Execute SQL queries against connected data sources and retrieve results. When using this tool, ensure you use the format catalog.schema.tableName for table names.
• execProcedure: Execute stored procedures against connected data sources.

Data Discovery

• getCatalogs: Retrieve a list of available connections from CData Connect AI. Connection names are used as catalog names in other tools and queries.
• getSchemas: Retrieve a list of available database schemas for a specific catalog. Use getTables to get tables within a schema.
• getTables: Retrieve a list of available database tables for a specific catalog and schema. Use getColumns to get columns for a specific table.
• getColumns: Retrieve a list of available database columns for a specific catalog, schema, and table.

Stored Procedures

• getProcedures: Retrieve a list of stored procedures for a specific catalog and schema.
• getProcedureParameters: Retrieve parameter metadata for a specific stored procedure (catalog, schema, and procedure name required).

Getting Started

Step 1: Configure Authentication

1. In Copilot Studio, add the CData Connect AI MCP Server connector
2. When prompted for credentials, enter:
   • Username: Your Connect AI email address
   • Password: Your Connect AI password

Step 2: Discover Your Data

Start by using discovery tools to understand your available data:

1. Use getCatalogs to see your connected data sources
2. Use getSchemas with a catalog name to see available schemas
3. Use getTables with catalog and schema to see available tables
4. Use getColumns with catalog, schema, and table to see column details

Step 3: Query Your Data

Execute queries using the queryData tool with proper three-part naming:

SELECT * FROM catalog.schema.tableName

Common Use Cases

• Cross-source analytics: Query data across multiple data sources in a single query
• Data discovery: Explore available tables, columns, and relationships through natural language
• Report generation: Execute complex queries and format results for reports
• Stored procedure execution: Run business logic stored in your data sources

API Documentation

For detailed information about CData Connect AI:

• Documentation: https://cloud.cdata.com/docs/MicrosoftCopilot-Client.html
• Integrations guide: https://cloud.cdata.com/docs/Integrations.html

Support

For technical support or questions:

• Email: support@cdata.com
• Website: https://www.cdata.com/support
• Documentation: https://cloud.cdata.com/docs/MicrosoftCopilot-Client.html

Known Limitations

• Requires active CData Connect AI subscription with at least one configured data source
• Personal Access Token must be valid and have appropriate permissions
• Query performance depends on underlying data source capabilities
• Table names must use three-part naming convention: catalog.schema.tableName

Version History

• 1.0.0: Initial production release with full MCP tool suite

Throttling Limits

Actions

CData Connect AI

OAuth Authorization Server Metadata