Agent in a day - Online workshop

At a glance

Note

Agent in a Day instructor-led training

Continue your learning journey with the Agent in a Day instructor-led workshop. This course provides hands-on experience for building agents! Browse training events.

In this learning path you'll complete hands on activities for building and customizing agents in Microsoft 365 Copilot Chat and Microsoft Copilot Studio. You'll build a declarative agent in Microsoft 365 Copilot Chat then move to Microsoft Copilot Studio to build and customize a custom engine agent. In Microsoft Copilot Studio, you'll build a Contoso support agent that utilizes agent knowledge, tools, agent flows, generative orchestration, and autonomous functionalities.

This Learning Path is designed for use with the Agent in a Day instructor-led course. For more information on attending this free, instructor-led course, see Microsoft Power Platform instructor-led training. On that page, you can Register for an event.

Prerequisites

  • You need an environment with a licensed user to complete these exercises.

Modules in this learning path

Microsoft 365 Copilot Chat makes it easy to build declarative agents that enhance productivity across your organization. In this module, you'll explore the fundamentals of agent creation using Microsoft 365 Copilot Chat with templates and starting from scratch for a customer service declarative agent.

In this module, you build a conversational agent in Microsoft Copilot Studio, starting with the importing an agent solution, building an agent using natural language, and preparing the agent with knowledge sources to ground its answers in real-world data.

In this module, you extend your agent's capabilities in Microsoft Copilot Studio by giving it access to tools, learn the difference between the different tools, build a custom prompt, and execute an order cancellation using an agent flow.

In this module, you extend your agent's capabilities in Microsoft Copilot Studio by giving it access to autonomous triggers, customize the agent's triggers to perform conditionally when specific conditions are met, learn how to test and validate your trigger's behavior, monitor the agent's activity, and publish changes to the agent.