Course
digicode: AB620
Design and Build Integrated AI Agent Solutions in Copilot Studio – Intensive Training (AB-620)
AB-620
Course facts
Download as PDF- Designing Copilot Studio agents that integrate with enterprise systems and data using connectors, APIs, Azure AI Search, and Model Context Protocol (MCP)
- Building and orchestrating advanced workflows and multi‑agent solutions
- Automating complex scenarios, including human‑in‑the‑loop and computer‑use workflows
- Applying identity, security, governance, and responsible AI considerations during solution design and implementation
- Testing, monitoring, and managing agents as enterprise‑grade solutions
You will go beyond creating standalone agents and learn how to deliver end‑to‑end solutions that automate real business processes, connect to enterprise data, and take meaningful action across platforms. Throughout the course, you will apply best practices for integration, governance, security, and responsible AI, so your agents are ready for real‑world deployment.
1 Deliver rich agent responses using Adaptive Cards in Microsoft Copilot Studio
Learn to configure rich agent responses in Microsoft Copilot Studio using message formatting and Adaptive Cards—covering informational and interactive card patterns, the built-in designer and JSON authoring, and channel-specific rendering considerations.
2 Take action from agent conversations using topics and tools in Microsoft Copilot Studio
Learn how to extend agent topics in Microsoft Copilot Studio to take action in external systems — calling tools, invoking agent flows, and configuring HTTP request nodes to retrieve and send data to external REST APIs.
3 Generate AI-powered agent responses using generative answers in Microsoft Copilot Studio
Learn how to configure AI-powered responses in Microsoft Copilot Studio agent topics. This module covers the generative answers node, connecting custom knowledge sources, using custom instructions to shape AI output, and authoring custom prompts with model selection from the Microsoft Foundry model catalog.
4 Design multi-agent solutions in Microsoft Copilot Studio
Learn to plan and design multi-agent solutions in Microsoft Copilot Studio. Evaluate when multi-agent approaches are appropriate, design solutions using child agents or connected agents, and select the right connection option and coordination pattern for your scenario.
5 Delegate agent tasks using child agents in Copilot Studio
Learn to structure complex Copilot Studio agents using child agents. Create focused subagents, configure how they receive tasks and return results, and control when they're invoked to build organized, scalable agent solutions.
6 Build multi-agent solutions using connected agents in Copilot Studio
Learn to build multi-agent solutions in Copilot Studio by connecting your orchestrator to existing agents built on Microsoft's agentic platform. Configure connections to Copilot Studio agents, Microsoft Foundry agents, and Microsoft Fabric Data agents, and manage your connected agent ecosystem for accurate orchestration.
7 Build cross-platform multi-agent solutions using the Agent2Agent protocol in Microsoft Copilot Studio
Learn to integrate cross-platform agents into your Copilot Studio solution using the Agent2Agent (A2A) protocol. Understand when A2A is the right integration choice, configure connections to external A2A-enabled agents with authentication, and validate end-to-end agent delegation.
8 Design integration strategies for agents in Microsoft Copilot Studio
Learn to evaluate and select integration patterns for your Copilot Studio agents. Explore tools, knowledge sources, and agents as integration options and identify appropriate authentication models to design integrations that meet your requirements.
9 Take action in external systems using connector and REST API agent tools in Microsoft Copilot Studio
Configure Power Platform connectors and REST APIs as tools for your Copilot Studio agents. Learn to add prebuilt and custom connectors, upload OpenAPI specifications to create REST API tools, and configure authentication to integrate your agent with external services.
10 Ground agents with enterprise knowledge using connectors and Azure AI Search in Microsoft Copilot Studio
Ground your Copilot Studio agents in enterprise knowledge by connecting to Copilot connectors, Power Platform connectors as real-time knowledge sources, and Azure AI Search vector indexes. Distinguish between index-based and real-time knowledge approaches and configure authentication settings for publishing knowledge-grounded agents.
11 Integrate agents with external systems via MCP in Microsoft Copilot Studio
Connect your agent to Model Context Protocol (MCP) servers to access standardized tools and resources. Learn to use the MCP onboarding wizard, configure API key and OAuth 2.0 authentication, manage tool selection, and determine when MCP is the right integration approach compared to connectors or REST APIs.
This course is designed for advanced builders and professional developers who want to design, extend, and integrate Copilot Studio agents. Your goal is to automate real business processes by connecting agents to enterprise data, systems, and services – while applying sound design, security, governance, and responsible AI practices.
You should already have hands‑on experience building agents in Copilot Studio, an intermediate understanding of generative AI concepts, proficiency working with REST APIs, and familiarity with the broader Microsoft cloud platform.
- Ability to create and publish an agent in Microsoft Copilot Studio
- Access to Microsoft Copilot Studio
- Basic familiarity with JSON
- Permissions to create solutions and agents in a Power Platform environment
This course can be used as part of your preparation for the «Microsoft Certified: AI Agent Builder Associate (beta)» exam.