GH-600 Course: Developing in Agentic AI Systems | GitHub and Agentic AI
The GH-600 course: Developing in Agentic AI Systems is an official Microsoft and GitHub training program aimed at professionals who want to design, deploy, and manage agent-based artificial intelligence systems within modern development environments. The training focuses on using GitHub as a reference platform to coordinate, monitor, and govern AI agents that actively participate in the software development lifecycle.
Agentic AI technology makes it possible to create autonomous agents capable of planning, reasoning, using tools, interacting with execution environments, and collaborating with other agents to solve complex tasks. Integrated with GitHub, GitHub Copilot, MCP (Model Context Protocol), and DevOps workflows, this technology helps accelerate development, improve team productivity, and automate software engineering processes.
In enterprise environments, Agentic AI systems are used to automate development tasks, code reviews, quality analysis, workflow execution, and the coordination of complex processes within the SDLC. The course provides practical insight into how to design secure, reliable, and governed agent architectures for use in production environments.
In addition, this course is official from Microsoft, prepares you for the GitHub Certified: Agentic AI Developer certification, and is eligible for FUNDAE subsidies for businesses.
Course overview
The GH-600 course: Developing in Agentic AI Systems provides the knowledge necessary to design, implement, and manage agent-based artificial intelligence systems within modern development environments using GitHub as a control and monitoring platform.
During the training, participants will learn how to integrate AI agents into the software development lifecycle (SDLC), define agent architectures, configure tools and execution environments, manage memory and state, coordinate multiple agents, and apply monitoring and governance mechanisms to ensure reliable and secure results.
The course also covers key aspects related to GitHub Copilot, MCP Servers, observability, result evaluation, autonomous behavior control, and protection through guardrails. The practical approach allows you to understand how to deploy AI agents in real production environments while maintaining the levels of quality, security, and control required by modern organizations.
Virtual course with certification exam included free of charge. Don't miss this opportunity! The exam is valued at €126 + VAT and is included at no additional cost.
Promotion valid until December 31, 2026. One-attempt exam available only in Virtual - Distance Learning format.
What is the Microsoft GH-600 certification used for in a professional environment?
The GH-600 certification: Developing in Agentic AI Systems accredits advanced skills in the design, deployment, monitoring, and governance of agent-based artificial intelligence systems within modern development environments. It is aimed at professionals who work with GitHub, GitHub Copilot, and Agentic AI architectures to automate complex processes in the software development lifecycle.
In the job market, this certification is especially relevant for roles such as AI Engineer, Developer, DevOps Engineer, Solution Architect, Platform Engineer, and professionals responsible for integrating artificial intelligence into development processes. Organizations are increasingly looking for specialists capable of coordinating autonomous agents, monitoring their behavior, implementing control mechanisms, and ensuring secure and reliable results in production environments.
Obtaining the GH-600 certification improves employability and facilitates specialization in one of the most innovative areas of artificial intelligence applied to software development. It also allows you to participate in projects where intelligent agents collaborate with technical teams to accelerate software delivery, improve productivity, and optimize engineering processes.
Professional applications of GitHub and Agentic AI
GitHub and Agentic AI represent a new generation of technologies aimed at the intelligent automation of software development. Through agents capable of planning, reasoning, executing tasks, and using tools autonomously, organizations can accelerate development processes and improve the productivity of their technical teams.
In enterprise environments, Agentic AI solutions are used to automate programming tasks, manage code reviews, coordinate DevOps workflows, execute quality analysis processes, generate technical documentation, and collaborate with developers throughout the software lifecycle. GitHub acts as a control and monitoring platform, allowing the implementation of governance, observability, and security mechanisms to ensure reliable results.
The GH-600 course allows you to apply these technologies in real-world scenarios by designing and managing systems based on intelligent agents. Participants will learn how to integrate GitHub Copilot, MCP Servers, multi-agent architectures, and control mechanisms that enable the development of secure, scalable artificial intelligence solutions ready for modern enterprise environments.
What Microsoft official courses and certifications at Nanfor include
The training includes Microsoft Learn material, presentations by expert instructors, authorized labs, specialized tutoring, and certification preparation.
Nanfor combines official Microsoft content with expert guidance, allowing for training focused on practical application in real business environments.
Learn all the components
Advantages of GH-600 training
Training in GH-600: Developing in Agentic AI Systems allows you to acquire advanced skills to design, deploy, and govern agent-based artificial intelligence systems within modern development environments using GitHub as a control platform.
- Learn how to integrate AI agents into the software development lifecycle (SDLC).
- Design secure, scalable Agentic AI architectures ready for production.
- Work with GitHub Copilot, MCP Servers, and advanced automation tools.
- Manage memory, state, and the execution of intelligent agents.
- Implement monitoring, observability, and autonomous behavior control mechanisms.
- Coordinate multi-agent systems to solve complex development processes.
- Apply governance, security, and accountability practices in AI systems.
- Specialize in one of the fastest-growing areas within artificial intelligence-based software engineering.
In addition, this training provides practical knowledge on how to combine artificial intelligence, automation, and DevOps to accelerate development processes and improve the productivity of engineering teams.
Prerequisites
To get the most out of the GH-600 course, prior experience in software development, DevOps processes, and GitHub-based workflows is recommended.
Knowledge of the following is also recommended:
- GitHub and GitHub Copilot.
- Software development lifecycle (SDLC).
- Git, repositories, branches, and pull requests.
- Basic concepts of artificial intelligence and agents.
- Quality, code review, and security practices.
- General CI/CD concepts.
The course is primarily aimed at developers, DevOps Engineers, AI Engineers, solution architects, and professionals working on building and operating systems based on intelligent agents.
Preparation for the Microsoft GH-600 certification
This GH-600 course: Developing in Agentic AI Systems prepares you for the official GitHub Certified: Agentic AI Developer certification, providing the knowledge necessary to design, implement, monitor, and govern Agentic AI systems using GitHub, GitHub Copilot, and advanced architectures based on intelligent agents.
⏱️
Course duration:
100 hours
🔑
Classroom access:
3 months
General course information
Who is the Microsoft GH-600 course aimed at?
The GH-600 course: Developing in Agentic AI Systems is aimed at technical professionals who want to design, deploy, monitor, and govern agent-based artificial intelligence systems within modern software development environments.
It is especially suitable for roles such as AI Engineer, Developer, DevOps Engineer, Solution Architect, Platform Engineer, and professionals who work with GitHub, GitHub Copilot, advanced automation, and architectures based on intelligent agents.
It is also aimed at development platform managers, software engineering teams, and artificial intelligence specialists who need to integrate, coordinate, and control autonomous agents within productive workflows while maintaining appropriate levels of security, quality, and governance.
GH-600 course objectives and what you will learn
By the end of the GH-600 course, you will be able to design, implement, and manage agent-based artificial intelligence systems within the software development lifecycle, using GitHub as the central platform for monitoring and controlling these systems.
- Understand the fundamentals of agent-based artificial intelligence (Agentic AI).
- Design agent architectures integrated into modern SDLC processes.
- Configure tools, permissions, and execution environments for intelligent agents.
- Work with GitHub Copilot, MCP Servers, and advanced automation mechanisms.
- Manage memory, state, and context for AI agents.
- Monitor autonomous behavior through controls and observability mechanisms.
- Evaluate and optimize results generated by Agentic AI systems.
- Coordinate multi-agent architectures to solve complex development tasks.
- Implement guardrails, security policies, and governance mechanisms.
- Prepare for the GitHub Certified: Agentic AI Developer certification.
Elements of the Microsoft Learn GH-600 collection
The official Microsoft Learn collection associated with the GH-600 course includes the following learning paths:
- Developing in Agentic AI Systems – Part 1 of 2
- Developing in Agentic AI Systems – Part 2 of 2
These learning paths allow you to acquire the knowledge necessary to design Agentic AI architectures, integrate agents into the software development lifecycle, manage execution environments, coordinate multi-agent systems, and apply security and governance practices using GitHub as a control platform.
GH-600 Course Content: Developing in Agentic AI Systems – Official Program
Unit 1. Fundamentals of Agentic AI in GitHub
- What is Agentic AI?
- Agentic AI vs. AI assistants
- Characteristics of Agentic AI
- Why Agentic AI is important
- The agent workflow
- Copilot Agent lifecycle
- The agent lifecycle: Plan → Act → Evaluate
- GitHub as a platform for agents
- What can AI agents do?
- Planning, reasoning, and action
- Review and approval of plans
- Defining goals for agents
- Risks of Agentic AI
- Human-in-the-Loop
- Agent governance in GitHub
Unit 2. Agent Architecture Design and SDLC Integration
- Main components of an AI agent
- Defining agent tasks and responsibilities
- Overview of agent architectures
- Selecting an agent architecture
- Structured agent planning
- Defining the agent execution scope
- Multi-agent architecture
- Communication between agents
- Connecting agents to external systems
- Memory and context
- Coordinating agent workflows
- Validation and approval of plans
- Security and governance
- Anti-patterns in agent architectures
- Architecture best practices
Unit 3. Tools, MCP, and Agent Execution Environments
- Why agents need tools
- Designing tools for agents
- Configuring permissions for agent tools
- Understanding MCP (Model Context Protocol)
- MCP implementation
- MCP configuration and governance
- Selecting an execution environment
- Where agent tasks are executed
- Agent execution context
- Designing execution environments
- Execution limits
- Agent actions in development workflows
- Securing agent execution
- Implementing secure tool integrations
Unit 4. Multi-Agent Systems and Orchestration
- Why use multi-agent systems?
- Defining agent responsibilities
- Multi-agent SDLC example
- Coordinating multi-agent workflows
- Multi-agent orchestration patterns
- Orchestration using GitHub workflows
- Isolating agent execution
- Coordination using artifacts
- Managing handoffs between agents
- Conflict detection and resolution
- Observability
- Reliability and recovery
- Recovery in multi-agent systems
- Common design errors
- Designing multi-agent systems for production
- Managing the agent lifecycle
Unit 5. Memory, State, and Evaluation
- Memory vs. state
- Designing agent memory
- Implementing memory strategies
- Managing the memory lifecycle
- Agent state persistence
- Capturing state using artifacts
- Managing context drift
- Maintaining continuity between systems
- Evaluation fundamentals
- Analyzing agent failures
- Designing evaluation metrics and signals
- Improving agent behavior
- Tuning and optimizing agent behavior
- Common memory and state issues
- Building reliable stateful agents
Unit 6. Governance, Guardrails, and Operations
- Risk-based autonomy
- Designing effective guardrails
- Implementing guardrails
- Governance controls in GitHub
- Designing Human-in-the-Loop workflows
- Implementing human approvals
- Balancing automation and human review
- Agent permission control
- Blocking of insecure actions
- Observability and auditability
- Agent operation in production
- Governance lifecycle
- Common governance errors
- Production readiness checklist
Our differentiator: Hands-on labs
One of the distinguishing features of the GH-600: Developing in Agentic AI Systems course is its practical approach focused on real-world scenarios for developing, deploying, and governing intelligent agents. Through guided labs, participants learn to work with GitHub, GitHub Copilot, Agentic AI architectures, and multi-agent systems used in modern software engineering environments.
The hands-on exercises allow participants to understand how to integrate agents into the software development life cycle (SDLC), monitor their behavior, manage tools and execution environments, and apply the control and governance mechanisms necessary for production environments.
| Nanfor Labs |
Technical skills developed |
Practical learning outcome |
| Lab 1: Fundamentals of Agentic AI in GitHub |
Understanding of Agentic AI architectures, how intelligent agents work, integration with GitHub, and use within the software development cycle |
Ability to identify scenarios where agents can provide automation and productivity within development teams |
| Lab 2: Agent architecture design |
Defining agent responsibilities, separation of planning, reasoning, and execution tasks, and design of interaction flows |
Building reliable, scalable Agentic AI architectures ready for enterprise environments |
| Lab 3: Agent integration in the SDLC |
Incorporating intelligent agents into development workflows, integration with repositories, code reviews, and DevOps processes |
Ability to automate tasks within the software lifecycle using AI agents |
| Lab 4: Tool and MCP Server configuration |
Configuration of external tools, Model Context Protocol (MCP), permissions, and service interaction capabilities |
Creating secure environments where agents can use controlled tools and resources |
| Lab 5: Memory, state, and context management |
Management of agent memory, context persistence, and state control during complex executions |
Implementation of agents capable of maintaining consistency and continuity in long-running processes |
| Lab 6: Secure execution of autonomous agents |
Execution control, establishment of operational boundaries, permission management, and behavior policies |
Ability to deploy agents with protection mechanisms that reduce operational risks |
| Lab 7: Agent monitoring and observability |
Monitoring of activities, event analysis, metric generation, and autonomous behavior tracking |
Complete visibility into agent operation and the ability to intervene when necessary |
| Lab 8: Evaluation and optimization of results |
Response validation, quality analysis, performance evaluation, and continuous agent improvement |
Ability to increase the accuracy, efficiency, and reliability of Agentic AI systems |
| Lab 9: Multi-agent system coordination |
Orchestration of multiple specialized agents, task coordination, and distribution of responsibilities |
Implementation of complex solutions where several agents collaborate to achieve shared goals |
| Lab 10: Governance, guardrails, and accountability |
Application of security policies, approval mechanisms, auditing, and intelligent agent governance |
Ability to deploy Agentic AI solutions aligned with enterprise security and compliance requirements |
Thanks to these labs, participants develop practical experience in building and operating Agentic AI systems, acquiring knowledge directly applicable to real projects involving the development, automation, and governance of intelligent agents on GitHub.
Course language
- Course: English / Spanish
- Labs: English / Spanish
Want to take this course? Request information now
If you want to take this training virtually, you can purchase it at the top of the product page. If you have any questions, please contact us.
If you want to take it in person or remotely, please contact us:
Nanfor, Microsoft's official IT training center
Nanfor is a custom IT training center specializing in technology training for professionals and businesses, and it is officially accredited by Microsoft as:
- Microsoft Solutions Partner – Training Services
- Microsoft Cloud Partner
These accreditations certify that Nanfor meets Microsoft's standards for delivering technical courses, using Microsoft content and Microsoft Certified Trainers (MCTs), ensuring quality, continuous updates, and alignment with certifications.
Frequently asked questions about the GH-600 course
What is the GH-600 GitHub and Agentic AI course?
The GH-600 GitHub and Agentic AI course is official training from Microsoft and GitHub focused on the design, deployment, and governance of agent-based artificial intelligence systems within modern software development environments.
What is the GH-600 GitHub and Agentic AI certification used for?
The GH-600 certification validates competencies in creating, monitoring, and managing Agentic AI systems, helping organizations integrate intelligent agents into software development, automation, and operation processes.
Is the GH-600 GitHub and Agentic AI course official from Microsoft?
Yes. The GH-600 course is official training from Microsoft and GitHub based on Microsoft Learn and aligned with the GitHub Certified: Agentic AI Developer certification.
How does the GH-600 GitHub and Agentic AI course differ from other Microsoft AI-related certifications?
The GH-600 course specializes in Agentic AI systems, covering agent architectures, multi-agent coordination, observability, MCP Servers, autonomous monitoring, and governance within GitHub and DevOps environments—areas that are not typically covered in more traditional AI certifications.
Is the GH-600 GitHub and Agentic AI course eligible for FUNDAE subsidies?
Yes. The GH-600 GitHub and Agentic AI course is eligible for FUNDAE subsidies for businesses, allowing companies to fully or partially recover their training investment depending on the available credits.
How long is the GH-600 GitHub and Agentic AI course?
The GH-600 course consists of 100 hours of training through Nanfor's virtual classroom, supplemented by official materials, hands-on labs, and additional learning resources.
Is the GH-600 GitHub and Agentic AI course included in Nanfor's LaaS?
Yes. The GH-600 GitHub and Agentic AI course can be part of Nanfor's LaaS programs, providing continuous access to specialized training in artificial intelligence, automation, and modern software development.
What technical level is required for the GH-600 GitHub and Agentic AI course?
The GH-600 course is aimed at an intermediate level. Previous experience with GitHub, GitHub Copilot, SDLC, DevOps, software development, and basic concepts of artificial intelligence and agents is recommended.
What career opportunities does the GH-600 GitHub and Agentic AI course offer?
The GH-600 course helps develop profiles such as AI Engineer, Developer, DevOps Engineer, Platform Engineer, Solution Architect, or specialist in intelligent agent-based automation within modern enterprise environments.
