Everyone is talking about AI agents. Most demos look impressive. Very few solve a real business problem that somebody can use tomorrow.

In this session I'll show how to build a simple but genuinely useful agent using Copilot Studio, MCP (Model Context Protocol), and Microsoft Dynamics 365 Business Central.

The scenario is straightforward: a customer sends an email asking for their account balance. The agent monitors the mailbox, retrieves the customer's information from Business Central through MCP, checks outstanding and overdue invoices, generates a response, and sends it back automatically.

Along the way we'll explore how MCP connects AI to business systems, how agents can securely work with business data, and why simple agents often deliver the fastest return on investment.

I'll also give an overview of the different options available for building agents around Business Central, where Copilot Studio fits in the picture, and how you can decide which approach is right for your organization.

No hype. No futuristic demos. Just a practical, end-to-end example that you can take back to work and start adapting to your own business processes immediately.

Key takeaways:
- Understanding MCP and why it matters
- Building a practical agent with Copilot Studio
- Connecting AI to Business Central data
- Automating customer interactions with email
- Exploring the growing Business Central agent ecosystem
- Identifying quick-win opportunities in your own organization