AB-620: Design and build integrated AI agent solutions in Copilot Studio

This intermediate-level course prepares learners to design, build, integrate, deploy, and manage production-ready AI agents using Microsoft Copilot Studio. Participants learn how to create sophisticated agent solutions that connect with enterprise systems, automate business processes, support multi-agent scenarios, and follow governance, security, and responsible AI best practices.

This course is designed for:

• Experienced Copilot Studio makers and professional developers.
• Technical professionals who want to build and extend AI agents for business processes.
• Developers who need to integrate agents with enterprise systems, APIs, and organisational data.
• Teams responsible for solution architecture, automation, governance, and AI implementation.
• Learners who already have practical experience with Copilot Studio and generative AI concepts.

Students should have:

• Hands-on experience building agents in Microsoft Copilot Studio.
• An intermediate understanding of generative AI concepts.
• Experience working with REST APIs.
• Familiarity with Microsoft cloud technologies and services.
• Basic understanding of artificial intelligence agents.
• Familiarity with Power Platform connectors, workflows, identity concepts, and governance principles.
• Access to Microsoft Copilot Studio.

• Basic familiarity with Markdown syntax

Licensing requirements: Microsoft Copilot Studio access is required for hands-on activities.

After completing this course, students will be able to:

• Design and configure AI agents in Microsoft Copilot Studio.
• Select appropriate agent harnesses and architectures for business scenarios.
• Control agent usage, capacity, and operational costs.
• Create instruction-driven agents with reusable skills and behaviours.
• Automate business processes using workflows and human approval patterns.
• Integrate agents with enterprise data, APIs, tools, and services.
• Connect agents to knowledge sources, MCP servers, and Microsoft IQ services.
• Design and implement multi-agent solutions.
• Apply governance, security, and responsible AI practices.
• Evaluate, monitor, and optimise agent performance in production.
• Implement application lifecycle management processes for agent solutions.
• Deploy scalable, enterprise-ready AI agent solutions.

1. Explore Microsoft Copilot Studio and agent harnesses

Learn how Microsoft Copilot Studio supports different agent runtime options and how harness selection impacts capabilities, orchestration, publishing, and billing. Participants compare harness types and select the most appropriate option for business requirements.

2. Manage usage and cost for agents in Microsoft Copilot Studio

Explore how Copilot Credits are consumed across agent types and learn how design decisions affect cost management, capacity planning, and monitoring.

3. Create instruction-driven agents in Microsoft Copilot Studio

Learn how to define agent purpose, behaviour, boundaries, grounding, and instructions using natural language, then test and refine agent performance.

4. Design reusable behavior with skills in Microsoft Copilot Studio

Create reusable skills that can be packaged and shared across agents and environments to improve consistency and maintainability.

5. Design integration strategies for agents in Microsoft Copilot Studio (GitHub Copilot)

Evaluate business requirements and identify the most suitable integration methods, access models, and capabilities for enterprise agent solutions.

6. Automate agent tasks with workflows in Microsoft Copilot Studio

Build deterministic workflows that enable agents to perform repeatable actions, integrate with systems, and deliver reliable business outcomes.

7. Add human-in-the-loop steps to workflows for Copilot Studio agents

Incorporate approvals, decisions, and information requests into workflows when human input is required during automated processes.

8. Ground agents in enterprise knowledge in Microsoft Copilot Studio (GitHub Copilot)

Connect agents to enterprise knowledge sources and validate that responses use trusted information effectively and accurately.

9. Extend agents with tools in Microsoft Copilot Studio (GitHub Copilot)

Enable agents to use tools that interact with enterprise systems, retrieve information, and perform authorised actions.

10. Connect agents to MCP servers in Microsoft Copilot Studio (GitHub Copilot)

Learn how to connect agents to Model Context Protocol (MCP) servers to access custom business services and internal systems securely.

11. Connect agents to Microsoft IQ in Microsoft Copilot Studio (GitHub Copilot)

Use Foundry IQ, Fabric IQ, and Work IQ services to provide agents with business knowledge, organisational context, and enterprise data.

12. Design multi-agent solutions in Microsoft Copilot Studio

Learn when multi-agent architectures are appropriate and how agents can coordinate work across complex business solutions.

13. Build connected agent solutions in Microsoft Copilot Studio (GitHub Copilot)

Create connected agent solutions that delegate tasks to specialist agents and support collaboration across workflows and business processes.

14. Design a governance and security strategy for agents in Microsoft Copilot Studio

Develop governance, authentication, security, and deployment strategies that support enterprise-scale agent solutions.

15. Design a responsible AI strategy for agents in Microsoft Copilot Studio

Apply responsible AI principles by implementing moderation controls, guardrails, transparency measures, and safety practices.

16. Evaluate agents in Microsoft Copilot Studio (GitHub Copilot)

Use structured evaluation methods to measure agent quality, test improvements, and validate model changes before deployment.

17. Monitor and improve agents in Microsoft Copilot Studio (GitHub Copilot)

Monitor production performance, investigate issues, analyse user interactions, and continuously improve agent effectiveness.

18. Implement application lifecycle management for agents in Microsoft Copilot Studio with Power Platform

Establish repeatable deployment and lifecycle management processes using solutions, pipelines, environments, and governance controls.