2025 Release Wave 1: Microsoft Copilot Studio and Microsoft Dataverse updates

Mar 31, 2025

sales, marketing and service
Microsoft

We're thrilled to introduce the first release wave of 2025 for Microsoft Copilot Studio and Microsoft Dataverse. This update is packed with new features and improvements designed to boost your business success. Continue reading to discover the most significant changes and updates for all your applications.

Microsoft Copilot Studio

Add suggested prompts to custom agents

Just like any standard Copilot, you can now add suggested prompts that your custom agent will support. This feature helps address issues with end users not discovering the full value of the Copilot extension and understanding its capabilities. These suggested prompts guide users in utilizing the agent effectively, ensuring they make the most of its features.

Use prebuilt agents as a starting point for Copilot creation

You can easily create and start running agents by using prebuilt agents as a starting point for Microsoft Copilot. The main aim is to accelerate the development process and utilize templates where best practices and design guidelines are incorporated from the start. These prebuilt agents help streamline the creation process, ensuring that your Copilot is built efficiently and effectively.

Get enhanced diagnostics to help troubleshoot your agent

This update brings key improvements to Microsoft Copilot Studio’s debugging tools by providing more detailed tracing and insights into the actions an agent took. These diagnostics include analysis of the steps taken by an agent to respond to a message, any errors encountered, and suggestions to resolve issues. The main aim is to reduce the time it takes to troubleshoot an agent. These enhancements help streamline the debugging process, making it easier to identify and resolve issues efficiently.

Analyze action and knowledge source usage for agents

This release wave will make it easier to understand and optimize how actions are used by your organization's agents. Enhanced visibility and analytics will help identify usage patterns, areas for improvement, and optimize performance.

Feature details

Comprehensive metrics: Metrics like answer rate, action usage rate, and error rate will provide a detailed view of user interactions. Detailed error breakdowns will help pinpoint and address specific issues effectively.
Contextual usage analytics: Understand how well a knowledge source is performing. Metrics introduced include knowledge source usage (measuring source popularity/effectiveness), answer rate (measuring source reliability/accuracy), and error rate (identifying and addressing issues with the source).

Use Microsoft Teams chats as knowledge sources

This improvement will be beneficial for organizations and teams that heavily rely on Microsoft Teams conversations. Copilot agents will be able to provide more accurate answers based on specific interactions within Microsoft Teams chats, preserving institutional knowledge.

Feature details

Adding Microsoft Teams knowledge sources: Makers can add Teams knowledge sources as a URL endpoint or through a data source selector, including messages from channels, meetings, and group chats.

Response accuracy: The agent will respond with information from Microsoft Teams message threads, depending on the maker's and end user's access to the specified chats, channels, or meetings.

Manage business rules for omnichannel agents via Microsoft Power Apps

This feature is particularly useful for enterprises where rapid changes to IVR (Interactive Voice Response) agents are required. This can be especially important for enterprises needing rapid changes in response to external factors such as natural disasters, power outages, or financial shifts like market fluctuations or policy amendments. Without needing to modify the underlying logic of an IVR agent, business administrators can adjust the way call flows are handled by the agent, providing greater flexibility in real-time and helping businesses maintain continuity.

Set-up steps:

Create and configure your Dataverse table in Microsoft Power Apps.
Create an application in Microsoft Power Apps to support the administration of your data.
Set up an action in Copilot Studio to connect your Dataverse table to your agent.

explore further on the Microsoft website

Microsoft Dataverse

Enhance AI-powered experiences with Microsoft Dataverse search

Microsoft Dataverse search will be updated with an improved indexing process, becoming a requirement for all generative AI-powered experiences, such as Copilot. Due to this dependency, the storage consumption of Dataverse search will increase to reflect the improved semantic Copilot indexing. Multiple features within the Power Platform are or will become dependent on Dataverse search.

Feature details

Monitor consumed storage: You will be able to monitor the consumed storage at the environment level through two tables:
- Dataverse search – Structured index: For database storage indexing.
- Dataverse search – Unstructured index: For files storage indexing.

Create reusable functions with Microsoft Power Fx

Functions in Microsoft Dataverse are reusable pieces of business logic that execute specific commands within Dataverse, running server-side. You will be able to call these functions from different places within the Microsoft Power Platform, such as Power Apps or within Power Automate flows.