Upcoming Sessions

Session with asterisk (*) are guaranteed to run

Dec 14 - 16, 2026 12:30pm - 5:00pm EST

Course Format and Delivery

Delivery Method: Instructor-Led Training (ILT or Instructed-Led Online (ILO)
Schedule: 3 sessions of 4.5 hours
Cost: $1495 USD

Group Discount: 20% off for 3+ attendees from the same organization.

About this course

AI agents don’t just respond. They plan, reason, use tools, and complete work. In this practical course, you will build five working agents on Microsoft’s 2026 agent platform. You’ll ground them in enterprise knowledge, connect them to tools and workflows, publish them to the web and Microsoft Teams, and learn how developers call agents in code with Microsoft Foundry. Governance, security and cost control are built into every module.

• Explain how AI agents work and when an agent beats rule-based automation
• Choose the right Microsoft tool for the job: Microsoft 365 Copilot, Copilot Studio or Microsoft Foundry
• Build and publish agents with instructions, topics, variables and enterprise knowledge
• Connect agents to tools, agent flows, APIs and Model Context Protocol (MCP) servers
• Create and call agents in code with Python, C# and REST
• Deploy, monitor and govern agents for identity, safety and cost

What you will need

• Familiarity with Microsoft 365 and basic cloud concepts
• A high-level understanding of generative AI (Microsoft AI-901 or equivalent is recommended)
• No coding is required for the labs. Python or C# experience is helpful for the code walkthrough

This course is great for

• Business, operations and IT leaders automating processes with AI agents
• Solution architects integrating agents into Microsoft 365, Power Platform and Azure
• Developers, makers and MLOps engineers working with Copilot Studio, Azure OpenAI or Foundry
• Product and innovation teams prototyping assistants, copilots and workflow tools

Topics Covered

DAY 1 • FOUNDATIONS AND PLATFORM

Module 1: Foundations of Agentic AI

• Conversational vs. generative vs. agentic AI
• LLMs, hallucination and retrieval-augmented generation (RAG)
• Azure OpenAI in Microsoft Foundry: models, quotas and pricing
• Agent anatomy and Microsoft’s 2026 agent stack

Module 2: Inside Copilot Studio and Microsoft Foundry

• Architecture, agent types and Microsoft Dataverse
• Harnesses, licensing and Copilot Credits
• Environments, solutions and governance

Lab 1:

Module 3: Creating Agents

• Templates, natural language and instructions
• Models, orchestration, settings and testing
• Publishing an agent

DAY 2 • BUILD AND DESIGN

Lab 2

Module 4: Topics and Conversation Design

• Triggers, nodes, questions and conditions
• Variables, entities and Adaptive Cards

Lab 3

Module 5: Knowledge and Grounding

• Files, websites, SharePoint and Dataverse as knowledge
• Citations, generative settings and publishing to Teams

DAY 3 · ACT, DEPLOY AND SCALE

Lab 4

Module 6: Tools, Actions and Agent Flows

• Connectors, agent flows, REST APIs and MCP
• Tool settings, approvals and credit costs

Lab 5

Module 7: Coding, Deploying and Monitoring Agents

• Microsoft Foundry Agent Service in Python, C# and REST
• Monitoring, evaluation, safety and cost controls

• Customer support, finance, healthcare and manufacturing use cases
• Multi-agent patterns, go-live and your capstone