Still manually translating requirements, arguing about fit/gap, writing X++ line by line, compiling, fixing errors, deploying, testing… and repeating until morale drops? Cute.

In this session, we’ll show how Dynamics 365 Finance & Operations development evolves when you bring in an MCP server and a squad of AI agents that actually do the work. A requirement lands in Azure DevOps, and instead of becoming your next headache, it gets picked up by an orchestrator that delegates everything—summarization, architecture decisions, fit/gap analysis, design, development, deployment, and even testing.

Watch AI agents dig through standard and custom D365FO code, challenge assumptions (yes, really), generate designs you can argue with, and then go ahead and implement them anyway—creating tables, classes, forms, and extensions via X++ without touching XML like it’s 2012. They compile, fix their own mistakes (must be nice), deploy to real environments, and validate the solution like a functional consultant who never gets tired.

If something breaks? No drama. The agents loop, fix, redeploy, retest—until everything works and you’re left wondering what exactly you’re supposed to do now.

Welcome to AI-powered ERP development, where your biggest role might just be approving things and pretending you always trusted the process.

Level: Advanced (you should be comfortable with X++, D365FO architecture, and ALM concepts)
Target audience: Solution Architects, Senior Developers, Tech Leads, and curious Functional Consultants who suspect developers are about to be replaced (partially… relax)
Format: Live demo + real scenario walkthrough (yes, things might break—that’s part of the charm)
Technical focus: MCP server integration, multi-agent orchestration, X++ metadata automation, AI-assisted design & development lifecycle
Prerequisites: Basic understanding of Azure DevOps, D365FO development model, and why “just a small change” is never small
Session duration: 45–60 minutes (depending on how rebellious the agents feel)
First delivery: Conference-ready, polished demo—but still evolving as fast as the tech itself
Takeaways:
How to orchestrate AI agents across the full ERP development lifecycle
What actually works today vs. what is still “conference magic”
Where this approach fits (and doesn’t) in real D365FO projects

Optional disclaimer: No developers were harmed during the making of this demo… but a few may feel slightly threatened.