Introduction
Organizations often maintain support guides that answer common employee questions. However, employees might not know where to find the appropriate guide or how to locate the information they need.
Microsoft Copilot Studio is a graphical, low-code environment for creating and managing agents. An agent can answer questions and help complete tasks for employees or customers by using the instructions and knowledge sources you provide.
Important
Access to Microsoft Copilot Studio and agent usage might require an applicable license or billing plan. Before beginning the exercise, confirm that you have permission to create and test agents in your organization's Copilot Studio environment. Review the Microsoft Copilot Studio product and pricing information for more information.
In this module, you create an agent in Microsoft Copilot Studio for a defined business scenario. You plan the agent, configure its instructions and knowledge, test its responses, and refine its behavior.
| Agent component | What it defines |
|---|---|
| Purpose and audience | The job the agent performs and who it supports |
| Instructions | How the agent should approach requests |
| Knowledge | The approved information the agent can use |
| Expected behavior | How the agent handles common, unclear, and unsupported requests |
| Validation criteria | How success is measured |
Example scenario for this module
Suppose you work on the IT support team at Contoso, an outdoor equipment company. Employees frequently contact your team with questions about approved software, device setup, password resets, and common technical issues. Many of these questions can be answered by using the company’s existing IT support guide.
Your team wants to create an agent that helps employees find answers from this approved guide. The agent should provide relevant information, remain within the scope of internal IT support, and direct employees to the IT help desk when it can't provide an appropriate answer.
Before making the agent available to employees, you need to define its purpose and boundaries, add the approved knowledge source, and test how the agent responds to different requests.
Learning objectives
In this module, you:
- Plan an agent by defining its purpose, audience, scope, approved knowledge, expected behavior, escalation approach, and validation criteria.
- Create and configure an agent by defining its identity, writing instructions, and connecting an approved knowledge source.
- Test and refine an agent by submitting representative requests, comparing its responses with validation criteria, and updating its configuration.
To get started, you'll define the planning decisions that guide the agent's purpose, knowledge, behavior, and validation approach.