The Data Provisioning server automates the provisioning and management of enterprise data assets by connecting your AI Agents to Data Marketplace.
This MCP server bridges the gap between data discovery and data utility by automating the access request process. It connects AI agents to Cloud Data Marketplace (CDMP), allowing them to initiate governed provisioning workflows for enterprise data assets. By standardizing the path from finding data to using data, this server ensures that data democratization is balanced with strict governance. It requires valid business justifications for every access request.
Key capabilities
This MCP server offers the following key capabilities:
•Automated access requests. Streamlines the process of requesting data by initiating formal workflows directly from the AI interface.
•Marketplace integration. Connects seamlessly with Data Marketplace to offer a shopping-like experience for governed data sets.
•Governance-led provisioning. Integrates with CDGC policies to ensure that all data access follows enterprise compliance standards.
Use cases
Use this MCP server to address the following use cases:
•Just-in-time data access. Allows users to ask queries like, "Get me access to the Q3 Sales dataset," and initiates the request immediately.
•Self-service analytics. Empowers business users to obtain the data they need for their BI tools without opening manual IT tickets.
•Compliance-first sourcing. Ensures that sensitive data is only provisioned after the proper approvals and justifications are recorded.
•Automated onboarding. Automatically requests access to necessary data bundles for new team members based on their roles.
Tools
The following table describes the tools available on this MCP server:
Tool
Description
checkout_data_order
Initiates a formal access request workflow within Data Marketplace, linking users to governed data assets after submitting a business justification.
list_data_collections
Retrieves a list of data collections with optional filtering by name or by search term and selected fields.
list_delivery_targets
Retrieves detailed information about a specific data collection. You can use the optional segments query parameter to include related segments. For example: segments=all
Note:
In AI Agent Engineering, the MCP server tools are called "actions."