Course Overview
Embark on a four-day, hands-on journey into the world of low-code agent development with Copilot Studio, crafted for Power Platform Markers, Software Engineers and Solution Architects ready to bring intelligent automation to life within Microsoft 365.
Your learning adventure begins with a clear introduction to agents and copilots—what they are, how they operate within the Microsoft 365 ecosystem, and the licensing options available. You’ll explore the essential building blocks of agents: knowledge and tools. Foundational concepts such as prompt engineering, semantic indexing, Retrieval Augmented Generation (RAG), and the Model Context Protocol (MCP) are explained in practical terms. Early on, you’ll learn how to plan and envision agent scenarios and configure your environment for productivity, including no-code approaches with SharePoint and Agent Builder.
The journey continues as you dive into the essentials of Copilot Studio itself. You’ll create and configure agents, manage knowledge and starter prompts, and work with topics, entities, and variables. Power FX is introduced for adding logic, while conversation design best practices and Adaptive Cards help you build engaging and effective interactions. You’ll also see how to publish your agents to channels such as Teams and other common chat applications, making your solutions widely accessible.
A major focus is placed on extending your agents with powerful tools. You’ll learn to use prompt actions to guide agent behavior, leverage AI Builder and managed agents for document processing, and connect to enterprise data using Connectors and Dataverse. The course covers consuming REST APIs for external integrations and using the Model Context Protocol to enhance agent capabilities. You’ll also automate visual tasks with Computer Use, enabling agents to handle a broader range of business needs.
As you progress, advanced topics come into play. You’ll build Autonomous Agents that can act on triggers, orchestrate multiple agents to collaborate on complex workflows, and apply deep reasoning models to solve challenging problems. Agent flows are used to automate tasks, voice support is added for natural interactions, and pro-code agents are integrated for scenarios that require custom logic or advanced features.
You’ll also learn how to design and implement business processes with agents—defining scenarios, mapping them to agent capabilities, designing conversation flows, and orchestrating multi-agent solutions to automate real business operations.
The final part of your journey focuses on publishing, monitoring, and maintaining your agents. You’ll work with the Power Platform Admin Center, manage environments and solutions, and deploy agents to a variety of channels, including common chat applications and the web. Application lifecycle management is covered in depth, including exporting and importing agents between environments, setting up deployment pipelines, and integrating with source control for robust versioning. You’ll use analytics and insights to monitor agent performance, ensure compliance and security with Microsoft Purview, and apply strategies to prevent agent sprawl.
By the end of this course, you’ll have a comprehensive, practical understanding of how to build, extend, monitor, and maintain low-code agents with Copilot Studio—equipped to deliver intelligent automation and real business value.
Who should attend
- Power Platform Makers
- Software Engineers for Agentic AI
- Solution Architects for Agentic AI
Prerequisites
- GitHub User Account
- Basic Knowledge of the Microsoft 365 Platform
Course Content
Module 1: Introduction to Copilot Extensibility & Agents
- Copilot & Agents Overview
- What is Agentic AI?
- Introduction to Agents and Copilots for Microsoft 365
- The Microsoft 365 platform & Copilot for Microsoft 365 Ecosystem
- Copilot & Copilot Studio Licensing Options (Full vs PAYG)
- Key Components of an Agent: Knowledge & Tools
- Prompt Engineering, Semantic Index & RAG
- Tools & Model Context Protocol (MCP)
- Microsoft Copilot Extensibility Fundamentals
- Envisioning Planning & Scenarios
- Productivity & No Code Agents (SharePoint, Agent Builder)
- Tools & Environment Configuration
Module 2: Copilot Studio Basics
- Overview Copilot Studio Agents in Copilot Studio
- Creating & Configuring Copilot Studio Agents
- Mastering Knowledge & Starter Prompts
- Topics, Entities and Variables
- Introduction to Power FX
- Conversation Design Best Practices
- Collecting Information using Adaptive Cards
- Publish Agents to Channels (Teams, Charts)
Module 3: Extending Copilot Studio Agents Tools
- Prompt Actions,
- AI Builder & Managed Agents: Document Processor
- Integrate Enterprise data using Connectors
- Using Dataverse & Integrating Data
- Consuming REST Api’s
- Integrating Model Context Protocol (MCP)
- Automating Visual Tasks using Computer Use
Module 4 Advanced Copilot Studio Agents
- Autonomous Agents & Triggers
- Multi-Agent Orchestration
- Use Deep Reasoning models for complex tasks
- Using Agent Flows to Automate Tasks
- Adding Voice Support to Agents
- Integrating Pro-Code Agents
Module 5: Designing & Implementing Business Processes
- Defining Scenarios & Envisioning
- Mapping Scenarios to Capabilities
- Designing Conversation Flows
- Generating Agents Instructions & Requirements
- Implementing Business Processes
Module 6: Publishing & Maintaining Copilot Studio Agents
- PowerPlatform Admin Center & Architecture
- Introduction to Power Platform Environments & Solutions
- Publishing and Deploying Copilot Studio Agents
- Using Advanced Channels (Whatsapp, Web, ...)
- Application Lifecycle Management for Copilot Studio Agents
- Exporting & Importing Agents between Environments
- Deployment Pipelines & Source Control integration
- Using Analytics And Insights
- Copilot Studio Compliance & Security
- Introduction to Microsoft Purview
- Preventing Agent Sprawl
English
German