Copilot in finance and operations apps provides generative AI capabilities in a natural language chat experience in the finance and operations client. Makers of finance and operations apps must be able to extend and customize Copilot to tailor the capabilities and experience to the needs of the organization. Extending Copilot can be done by creating AI plugins that add needed capabilities.

With this release, we will expand the maker's ability to customize the Copilot experience by enabling the creation of AI plugins that are based in finance and operations business logic and code. Copilot can respond to prompts with plugins that run finance and operations methods in code to determine the answer or action in response to the prompt. These are headless actions that don't require a specific context in the finance and operations client. These actions can be registered in the Dataverse plugin registry and made available in any Copilot connected to the registry, like Copilot for Microsoft 365, other Copilot experiences for Dynamics 365 applications, or custom copilots.

These capabilities open up finance and operations business logic to be used in expanding the capabilities of Copilot in finance and operations in-app experiences, as well as across your Copilot ecosystem.

The Model Context Protocol is an open standard that facilitates the connection of AI agents to various systems, standardizing how the systems and applications provide context to large language models (LLMs) in agent platforms. With this release we will make available an MCP server for Dynamics 365 finance and operations apps, including key actions for Dynamics 365 Finance and Dynamics 365 Supply Chain Management as tools on the MCP server. This creates a streamlined experience for connecting any agent platform that supports the standard protocol, including Microsoft Copilot Studio, to the actions and business logic of Dynamics 365 finance and operations apps.

Copilot in finance and operations apps provides generative AI capabilities in a natural language chat experience in the finance and operations client. Makers of finance and operations apps must be able to extend and customize Copilot to tailor the capabilities and experience to the needs of the organization. Extending Copilot can be done by adding client actions, which are contextual copilot operations that can be completed in the flow of a business process in the finance and operations client.

With this release, the maker's ability is expanded to customize the Copilot experience by enabling the creation of client actions that are based in finance and operations business logic and code. Copilot can respond to prompts in the in-app chat experience with actions that run finance and operations method in code to determine the response. These capabilities, in conjunction with AI actions for finance and operations, open up finance and operations business logic to be used in expanding the capabilities of Copilot in finance and operations in-app experiences.