AI Agents in Action – Practical Use with Microsoft Copilot

2 x Half-Day Practical Training Sessions

Hands-on training to design, build, and deploy AI agents for workflow automation and real-world productivity Also known as Copilot Advanced Agents, and the third level of the Copilot progression, after Copilot Essentials and Copilot in Practice.

Summary:

A hands-on course focused on designing, building, and deploying AI agents using Microsoft Copilot to streamline workflows and boost real-world productivity.

Course Breakdown:

Building AI agents in Microsoft Copilot, from agents inside Copilot chat in session one to a full Copilot Studio build in session two.

Foundations

• Introduction to AI agents and Microsoft Copilot capabilities
• How Copilot agents work and where to find them
• Built-in Microsoft agents
• Copilot Agents vs Copilot Studio Agents, and how to decide between them

Workflow & Prompting

• Workflow audit and automation design fundamentals
• Creating effective prompts for business tasks
• Tips for prompting: the Prompt Coach

Building Your Own Agents, Session One

• Creating your own agent
• Configuring it
• AI Prompting Agent
• Business Communication Coach
• HR Policy Assistant Agent
• Simulator Agent
• Mission Expense Categorisation Agent
• Tender Review Agent
• Personal Data Spotter

Governance & Security

• Compliance, security and governance essentials
• Security reminder: safety, quality and trust

Copilot Studio, Session Two

• Where to find Copilot Studio
• Agent naming
• Model selection
• Instructions and testing
• Knowledge bases and RAG: retrieval-augmented generation
• Control layers: unstructured documents
• Structured FAQs
• Topic-based answers for high-risk policies
• Topics and nodes
• Leading a conversation
• Giving powers to your agent: tools and flows
• Power Automate
• MCP servers
• Tools and connectors: configuration and connections
• User authentication
• Building a topic that uses a tool, with escalation by email
• The finished agent
• What more could be done
• The limitations of Copilot Studio

Outcome:

Participants will gain practical experience with the latest AI technologies and leave with a clear understanding of how to apply and prepare for the next wave of AI innovation.

Participants will be able to build, test and govern their own Copilot agents, starting with chat-based agents they can create the same day and progressing to a fully configured Copilot Studio agent with its own knowledge base, controlled topics and working escalation. They will also leave knowing where the platform’s limits are and when an agent is the wrong answer.