Announcing new Microsoft Dataverse capabilities for multi-agent operations

Microsoft Dataverse continues to solidify its role as a pivotal platform for organizations seeking to harness the full potential of their agent ecosystems. As Microsoft Build 2025 unfolds, the spotlight shines on Microsoft Copilot Studio, which has unveiled a suite of innovative features that enhance agent capabilities. At the heart of these advancements lies Microsoft Dataverse, a secure and scalable platform that integrates enterprise data with powerful agent functionalities.

Dataverse: the operational database for agents

Dataverse serves as the backbone for organizations to store, manage, and orchestrate both business and operational data across their agent ecosystems. With Dataverse powering Microsoft Copilot Studio, developers can deploy agents capable of executing logic-driven, adaptive tasks at scale while ensuring necessary human oversight.

This platform fosters a cohesive environment where generative AI can be applied seamlessly to business and operational data. Features such as AI-powered Dataverse search, prompt columns, and the Model Context Protocol facilitate connected operations, allowing for a more integrated approach to data management.

For instance, the introduction of prompt columns enables users to embed generative AI directly into Dataverse tables. By utilizing a prompt column for “Sentiment” in a product review table, agents can automatically evaluate the review text, returning values like “Positive” or “Negative.” This dynamic business logic, powered by natural language prompts, exemplifies how Dataverse is laying the foundation for next-generation, business-aware agent systems.

Dataverse Model Context Protocol server

The Dataverse Model Context Protocol (MCP) server, now in public preview, transforms structured Dataverse information into interactive knowledge for Copilot Studio agents. This development is crucial for both developers crafting advanced workflows and makers configuring intelligent experiences, as it allows data to be conversational and accessible.

Once integrated, the MCP server unlocks four essential capabilities:

  1. Query: Discover available tables, explore schema, and retrieve real-time data through structured or natural language queries.
  2. Knowledge and search: Empower agents to engage with data, search knowledge sources, and provide contextual answers without cumbersome configurations.
  3. Upload (Create/Update Records): Insert new records or update existing ones in Dataverse while ensuring data integrity through schema-aware mapping.
  4. Generate with grounding prompts: Execute custom prompts that are rooted in real business contexts, such as summarizing records or evaluating sentiment.

By exposing the Dataverse environment through the MCP server, organizations can breathe life into their enterprise data, enabling agents to reason across structured data and generate meaningful outputs while adhering to established data models and access controls.

Enterprise data integration across Copilot Studio, Microsoft Fabric, and Microsoft 365 Copilot

Dataverse knowledge in Copilot Studio, now generally available

As agents rely heavily on the knowledge they can access, the AI-powered Dataverse remains integral to the Dataverse knowledge within Copilot Studio. This feature connects structured and unstructured data from various sources—including Dynamics 365, Power Platform, and external systems—creating a unified knowledge network that agents can leverage for informed decision-making. Recent enhancements include support for multi-line text and file type columns.

Use near-real-time data warehousing with Microsoft Fabric

All data housed in Dataverse is pre-indexed, facilitating near-real-time analytics. Whether it involves critical business data from Dynamics 365 or custom Power Apps applications, everything is securely stored in Dataverse and readily available for analysis. The integration with Microsoft Fabric allows users to explore this data effortlessly, unlocking profound insights.

Dynamics 365 data now available in Microsoft 365 Copilot

With the new integration, customers can now search and reason over Dynamics 365 data—such as contacts, accounts, leads, opportunities, and cases—directly within Microsoft 365 Copilot. This feature streamlines workflows by allowing users to glean insights and take action without the need to switch contexts.

Knowledge tools to retrieve the right context at the right time

The knowledge platform in Microsoft Copilot Studio, powered by Dataverse, enables customers to utilize their own data—such as local files and Dataverse tables—as knowledge sources for building intelligent, context-aware agents. This capability simplifies the process of unifying and operationalizing knowledge across the agent ecosystem.

New knowledge sources and connectors

To enhance agent capabilities, new knowledge sources have been introduced, including Snowflake, SAP, Databricks, Confluence (cloud only), OneDrive for Business, and SharePoint Lists. Support for Salesforce, Zendesk, and ServiceNow has also been expanded to include unstructured content, such as knowledge base articles. Additionally, Dataverse now supports image extraction, multilingual content, and querying embedded tabular data, further enriching the knowledge landscape.

Enhanced Power Platform connector Software Development Kit (SDK)

The newly enhanced Power Platform connector SDK, currently in preview, simplifies the integration of structured external data into Power Apps, Dataverse, and Microsoft Copilot Studio. This SDK allows connectors to expose structured data, enabling makers to bind tables directly to user interface controls in Power Apps and ground Copilot agents with business data.

Centralized Tools hub in Copilot Studio

The introduction of a centralized Tools tab in Copilot Studio provides makers with a dedicated space to create and manage reusable functionalities across all agents within an environment. This hub includes six tool types, such as the Model Context Protocol for connecting to existing knowledge servers and custom connectors for third-party systems.

New autonomous agents to put to work faster

To expedite the deployment of autonomous agents, Microsoft has introduced three new managed agents available in preview. The Document Processor Agent automates document workflows by monitoring email inboxes for attachments, extracting key information, and exporting structured data. This agent significantly reduces setup time and enhances efficiency.

  • The Customer Brief Agent generates timely executive briefs from business data before client meetings.
  • The Lead Manager Agent autonomously processes and responds to inbound leads.

These agents are designed to facilitate quick implementation and scalability, allowing organizations to harness the power of intelligent automation effectively.

Winsage
Announcing new Microsoft Dataverse capabilities for multi-agent operations