Microsoft Unleashes Dynamic AI Capabilities in Dynamics 365 ERP MCP Server, Empowering Partners and Custom Agents

The Dynamics 365 ERP Model Context Protocol (MCP) server plays a crucial role in linking AI with ERP systems. “MCP provides a unified framework, instead of relying on custom APIs or point-to-point integrations,” said Jared Hall, Principal Program Manager, Microsoft, in a company blog post. “MCP standardizes access to ERP operations, helping ensure consistency, context, and control.”

In March of this year, my colleague Tom Smith reported on the rollout of the Dynamics 365 ERP MCP server, along with the Dataverse MCP and Dynamics 365 CRM MCP servers. He noted the significant steps Microsoft had taken, with these deployments, toward agent interoperability.

Now, Microsoft has turned those steps into leaps, announcing a series of enhancements to the Dynamics 365 ERP MCP server, boosting its already formidable capabilities to standardize AI agents, apps, and services for accessing ERP data and performing tasks in D365, and to drive partner innovation.

Shifting Gears: From Static to Dynamic

The D365 ERP MCP server was initially launched as a static framework that included 13 out-of-the-box tools for D365 Finance and Supply Chain Management. This static approach focused on delivering essential ERP operations, which were particularly useful for guiding users new to the agentic AI space. However, these tools were fixed in their capabilities.

Now, Microsoft has introduced the dynamic ERP MCP server (currently in public preview) to address the growing demand for flexibility in developing more complex and customized AI applications and workflows. Developers are no longer constrained. They can now build custom agents freely to perform tasks that interact with the ERP system in the same way as a human would.

With the new server tools, agents leverage server forms to complete actions. “This evolution unlocks hundreds of thousands of ERP functions across tens of thousands of forms,” said Hall. “These forms — in Finance, Supply Chain, Human Resources, and Project Operations — are instantly accessible through MCP.”

Dynamics 365 ERP MCP Server for Analytics 

Microsoft’s D365 ERP MCP server for analytics applies the model-context approach to enhancing data insights through governed access to ERP analytics data, such as measures, dimensions, reports, and semantic models, in Business Performance Analytics (BPA). Hall describes the new server as “a bridge between ERP transactions and business intelligence.”

The standout benefit here is consistency — the ability to ensure that agents and users collaborating are using the same definitions. This is a critical requirement for building trust and scaling agents, since ERP data is not only accessible and available but also has clear, understandable, and consistent definitions.

Another Win for the Partner Ecosystem

Like many Microsoft innovations, the introduction of the two new servers has opened up opportunities for independent software vendors (ISVs) and system integrators, enabling them to build and publish flexible, secure ERP solutions and deliver informed applications based on sound analytics. Early examples include:

• Shop Floor by RSM reduces downtime and improves quality in manufacturing by enabling real-time issue resolution, process automation, and collaboration.
• PayFlow Agent by HSO enables more efficient invoice payment in accounts payable through automated payment inquiries and real-time status updates.
• Inbound Load Agent by Fellowmind converts emailed delivery notes into inbound loads in Dynamics 365, reducing manual errors by eliminating repetitive data entry.
• Quality Impact Recall Agent by Cegeka pinpoints product quality issues and maps their impact across inventory and customers.
• AI agents from Crowe handle a variety of unstructured, high-touch processes that can’t be automated with traditional techniques, enabling agentic data entry.
• Supplier Performance Insight Agent by KPMG automates finance workflows by integrating a company’s ERP data with external market signals to provide a holistic understanding of supplier reliability and risk.
• Autonomous AI agents from Annata automate and accelerate intricate service processes to secure parts, create appointments, and direct the entire service chain.
• Axtension Downtime Agentby SignUp Software suggests resolutions and generates alternative production plans in the event of machine failure in production environments.

Closing Thoughts

The new servers integrate custom agent functionality, data, and analytics, allowing organizations to accelerate development in AI-driven ERP systems. Once again, Microsoft has demonstrated its expertise in timing, not overwhelming users by introducing advanced functionality only when demand arises. This approach truly guides customers through the often confusing and crowded landscape of the AI Revolution.

And Microsoft is uniquely positioned to do this. With such a comprehensive product portfolio, they can roll out a wide range of products horizontally without scaling vertically too quickly.

Microsoft is making significant strides across its portfolio, but it doesn’t feel too much for customers because it’s incremental and, most importantly, responsive to user demand. This approach is thoughtful and considerate, benefiting Microsoft, its customers, and, as demonstrated by the variety of partner agents, the partner ecosystem as well.

---