Target Audience
This course is designed for advanced builders and professional developers who want to design, extend, and integrate Copilot Studio agents for enterprise-grade solutions. It is especially relevant for IT application developers, consultants, independent software vendor partners, app makers, Power Platform developers, and technical professionals responsible for building scalable AI solutions for organizations or customers. Participants should already be able to create, configure, test, and publish agents in Microsoft Copilot Studio. They should also have intermediate knowledge of generative AI, prompt engineering, REST APIs, integration patterns, Power Platform environments, Dataverse, Microsoft 365 Copilot, Microsoft Foundry, adaptive cards, and agent configuration with knowledge sources, instructions, tools, and topics.
Instructor-Led (Live Virtual/Classroom)
A live course led by a certified instructor, in virtual or classroom format — one full day of guided sessions with projects and case studies.
Course Overview
In AB-620T00-A: Design and build integrated AI agent solutions in Copilot Studio, learners develop the skills required to implement production-ready agents and multi-agent solutions capable of sophisticated reasoning, workflow automation, and enterprise integration. The course emphasizes agent architecture, advanced behaviors, external system integration, testing, deployment, monitoring, and lifecycle management. The official exam alignment focuses on three major domains: Plan and configure agent solutions, Integrate and extend agents in Copilot Studio, and Test and manage agents. Microsoft’s study guide assigns the largest weighting to integration and extension, followed by solution planning and agent management.
Last Updated: 2026-09-01
Course Exam Overview
- Program Name: none
- Included: Course, applicable labs, and virtual instructor-led facilitation included; certification exam attempt is managed separately when required
- Duration: Microsoft states that candidates have 120 minutes to complete this assessment
- Prerequisites: Candidates should have hands-on experience building agents in Copilot Studio, an intermediate understanding of generative AI concepts, proficiency working with REST APIs, and familiarity with the broader Microsoft cloud platform.
- Exam Format: Proctored Microsoft certification exam; Microsoft notes that the exam may include interactive components
- Delivery: Scheduled through Pearson VUE
- Outcome: Successful candidates earn the AI Agent Builder Associate certification, validating their ability to develop and integrate advanced, scalable AI agents using Microsoft Copilot Studio, Power Platform, and enterprise technologies.
Skills You’ll Gain
Job Roles & Industry Outlook
Copilot Studio Developer
Designs, builds, extends, tests, and deploys advanced Copilot Studio agents connected to enterprise systems and services.
AI Agent Builder
Creates production-ready AI agents, multi-agent solutions, advanced topics, tools, knowledge integrations, and secure business workflows.
AI Agent Builder
Builds integrated solutions using Copilot Studio, Power Platform, Dataverse, connectors, APIs, ALM practices, and enterprise governance.
Course Includes
• Three-day instructor-led training experience aligned to Microsoft Learn AB-620T00-A.• Technical guidance for building production-ready Copilot Studio agents and multi-agent solutions.
• Practical coverage of topics, tools, connectors, REST APIs, enterprise knowledge sources, Azure AI Search, Microsoft Foundry, and Microsoft Fabric Data agents.
• Applied exercises for agent design, integration, testing, monitoring, deployment, and application lifecycle management.
• Preparation alignment for the AI Agent Builder Associate certification exam, with exam attempt managed separately when required.
What You'll Learn
This module introduces the planning foundation for enterprise-grade Copilot Studio agent solutions. Learners evaluate business requirements, define solution scope, plan integrations, select identity and deployment approaches, and identify governance, security, and responsible AI considerations before implementation.
Lessons:
• Plan integration with enterprise systems
• Plan identity strategy for agent solutions
• Plan channels and deployment options
• Plan responsible AI strategy
• Evaluate security and governance considerations
• Plan reusable agent components
• Design agents for internal or external audiences
Key Topics:
• Agent solution architecture
• Enterprise integration planning
• Identity and access strategy
• Deployment channels
• Responsible AI and governance
• Reusable components
Labs / Practical Exercises (if applicable):
• Draft an agent solution design for an enterprise scenario
• Identify integration, identity, governance, and deployment requirements for a Copilot Studio agent
This module focuses on creating structured and intelligent agent conversations in Microsoft Copilot Studio. Learners configure topics, responses, variables, tools, generative answers, custom prompts, knowledge sources, and adaptive cards to deliver controlled, useful, and context-aware agent experiences.
Lessons:
• Configure topics in Copilot Studio
• Add agent flows to a topic
• Configure response formatting
• Add tools to a topic
• Configure advanced responses with custom prompts
• Configure advanced responses with custom knowledge sources
• Configure advanced responses with APIs and Send HTTP requests
• Configure generative answers nodes
• Configure adaptive cards
• Manage variables
Key Topics:
• Topic design
• Agent response formatting
• Custom prompts
• Custom knowledge sources
• Generative answers
• Adaptive cards
• Variables and tools
Labs / Practical Exercises (if applicable):
• Configure an advanced topic with formatted responses and variables
• Add a tool, custom prompt, or knowledge source to improve agent behavior
• Design an adaptive card interaction for a business scenario
This module covers agent flows as the workflow automation layer for Copilot Studio agents. Learners create flows, configure connectors and actions, define input and output parameters, implement human-in-the-loop patterns, monitor execution, and handle errors for reliable business process automation.
Lessons:
• Create an agent flow
• Create a human-in-the-loop agent flow
• Configure actions and connectors
• Add input and output parameters
• Monitor agent flows
• Implement error handling in agent flows
Key Topics:
• Agent flows
• Workflow automation
• Connectors and actions
• Human-in-the-loop design
• Parameters
• Monitoring and error handling
Labs / Practical Exercises (if applicable):
• Build an agent flow connected to a business process
• Add input and output parameters to pass data between an agent and a flow
• Configure basic error handling and review flow execution
This module develops the integration skills required to connect Copilot Studio agents with enterprise data, systems, and services. Learners connect to enterprise knowledge sources, use Copilot connectors and Power Platform connectors, integrate Azure AI Search, add REST APIs, configure MCP tools, and extend agents with custom connectors.
Lessons:
• Connect to Copilot connectors
• Connect to Microsoft Power Platform connectors
• Connect to Azure AI Search
• Configure and monitor computer use for an agent
• Configure MCP tools
• Add a tool using an existing custom connector
• Add REST APIs to an agent
Key Topics:
• Enterprise knowledge sources
• Copilot connectors
• Power Platform connectors
• Azure AI Search
• MCP tools
• REST APIs
• Custom connectors
• Computer use
Labs / Practical Exercises (if applicable):
• Map an agent to enterprise knowledge and system integration requirements
• Configure a connector-based tool for an agent
• Design an API integration pattern for a Copilot Studio agent
This module focuses on designing and implementing multi-agent architectures in Copilot Studio. Learners evaluate when multi-agent approaches are appropriate, design orchestration patterns, connect existing agents, integrate Microsoft Foundry agents, use Fabric Data agents, and create multi-agent solutions using A2A protocol.
Lessons:
• Design multi-agent solutions in Copilot Studio
• Integrate an existing agent in Copilot Studio
• Integrate a Foundry agent
• Integrate a Fabric Data agent
• Create a multi-agent solution using A2A protocol
• Select coordination and delegation patterns
Key Topics:
• Multi-agent architecture
• Agent orchestration
• Connected agents
• Microsoft Foundry agents
• Microsoft Fabric Data agents
• A2A protocol
• Delegation patterns
Labs / Practical Exercises (if applicable):
• Design a multi-agent architecture for a cross-functional business process
• Connect an existing Copilot Studio agent into a multi-agent solution
• Compare child-agent, connected-agent, and cross-platform agent patterns
This module addresses Azure and Microsoft Foundry integrations for advanced agent scenarios. Learners configure generative answers with Azure AI Search and Foundry, use the Foundry model catalog with custom prompts, and monitor agent behavior and performance through Application Insights.
Lessons:
• Configure generative answers using Azure AI Search with Foundry
• Configure custom prompts to use the Foundry model catalog
• Monitor agents using Application Insights
• Select Azure services to support advanced agent scenarios
• Align Azure integrations with security and governance requirements
Key Topics:
• Azure AI Search
• Microsoft Foundry
• Foundry model catalog
• Custom prompts
• Application Insights
• Monitoring and observability
Labs / Practical Exercises (if applicable):
• Design an Azure AI Search integration for generative answers
• Select a Foundry model approach for an advanced prompt scenario
• Review monitoring requirements for a production-ready agent
This module focuses on evaluating whether agents perform reliably and meet business expectations. Learners create test sets, choose evaluation methods, review test results, identify improvement areas, and apply structured validation before deployment or scale.
Lessons:
• Create a test set
• Choose an evaluation method
• Review test results
• Identify performance gaps
• Improve agent responses based on evaluation outcomes
Key Topics:
• Agent testing
• Test sets
• Evaluation methods
• Response quality
• Performance review
• Iterative improvement
Labs / Practical Exercises (if applicable):
• Create a test set for a Copilot Studio agent
• Evaluate agent responses against expected outcomes
• Document improvements based on test results
This module introduces application lifecycle management practices for Copilot Studio agents. Learners package agents into solutions, add existing agents to solutions, use environment variables, and implement or extend Microsoft Power Platform Pipelines to support controlled deployment across environments.
Lessons:
• Create a solution
• Add existing agents to a solution
• Create and use environment variables
• Implement Microsoft Power Platform Pipelines
• Extend pipelines for agent deployment
• Manage deployment readiness across environments
Key Topics:
• Application lifecycle management
• Solutions
• Environment variables
• Power Platform Pipelines
• Deployment governance
• Environment strategy
Labs / Practical Exercises (if applicable):
• Add a Copilot Studio agent to a solution
• Configure environment variables for deployment portability
• Outline an ALM process for moving agents across development, test, and production environments
Recommended Courses
Frequently Asked Questions
If you see Register Now, the course is free and available for direct registration; simply complete the form to secure your seat.
If you see More Information, share your contact details and an ITG Learning Advisor will contact you to confirm pricing, available dates, delivery options, and next steps.
We’ll help you find the right fit. Select More Information, and one of our ITG Learning Advisors will present other available dates, related courses, delivery formats, and learning paths based on your goals.
BlockSkill is ITG’s learning and skilling platform. It brings your learning experience together in one place, giving you access to your enrolled courses, learning materials, available videos, activities, surveys, progress information, and certificates of completion.
Available resources may vary depending on the course and delivery format.
An exam voucher is a code or authorization that covers one attempt at an official certification exam. A practice assessment is included only when it is specifically listed as part of the selected package.
To earn the official certification, you must pass the corresponding vendor exam or assessment. Ask an ITG Learning Advisor whether the exam voucher is included with your course.
Yes. You may purchase an exam voucher without enrolling in the related course, subject to vendor availability, exam eligibility, and regional conditions.
Our team can help you identify the correct voucher for your certification goal and explain the available purchasing options.
ITG offers private and customized training for teams, businesses, government agencies, and other organizations. We can help you select the right courses and coordinate schedules, delivery formats, and learning paths aligned with your organization’s goals.
Select More Information to connect with an ITG Learning Advisor and explore the best training option for your team
Ready to start?
Get certified.
Join the next live cohort of AI+ Everyone™ and earn your industry-recognized credential.
