Nick Perkins

Platform Engineer. Volunteer Motorsport Official. ADHD Brain. Bit of a nerd.

How I Built Interactive AI-Powered Training with GitHub Copilot

One of my current projects is a migration from Azure DevOps to GitHub Actions. I saw a challenge: how to effectively train DevOps teams on new tools in an engaging, practical way? Traditional docs and tutorials often miss the mark. So, I experimented: using GitHub Copilot’s Agent Mode to build an interactive, AI-powered training platform. The result was a self-paced, conversational learning experience, guiding users through complex migration scenarios like having an AI tutor.

Making Technical Training Engaging with AI

DevOps engineers need practical, interactive, adaptive, and contextual training. Traditional methods often fall short, unable to adapt to individual learning styles or keep pace with rapidly changing tech. To address this, I used GitHub Copilot’s Agent Mode to create an intelligent tutor for personalised, interactive learning. The platform has three key parts:

  • Topic-Based Folder Structure - Numbered folders with README.mds provide a clear, lightweight learning path. Folders allow you to add additional resources and examples where relevant.

  • AI Instruction Files - Each topic has an instruction file telling Copilot how to train, including pedagogical instructions, examples, exercises, assessments, and troubleshooting.

  • Global Copilot Configuration - A master instruction file (.github/copilot-instructions.md) defines the AI tutor’s overall behaviour: personality, learning methodology, audience understanding, and the “start the training” trigger.

The “Magic Trigger” is the secret sauce. Typing “start the training” into the GitHub Copilot chat while in a topic’s README makes Copilot identify the topic, load instructions, and begin an interactive, personalised session, adapting to user responses and providing real-time feedback.

How It Works in Practice

Here’s a typical learning session:

Starting a Session

# User opens: 08-migrating_nodejs_builds/README.md
# User types: "start the training"

AI Response

GitHub Copilot immediately transforms into a knowledgeable tutor:

🚀 Welcome to Node.js Build Migration!

I’m going to help you understand how to migrate Azure DevOps pipelines for Node.js applications to GitHub Actions.

Let me start by showing you a typical Azure DevOps pipeline, then we’ll convert it together step by step.

Here’s an Azure DevOps YAML pipeline for a Node.js app:

trigger:
- main

pool:
  vmImage: 'ubuntu-latest'

steps:
- task: NodeTool@0
  inputs:
    versionSpec: '18.x'

Before we convert this, tell me: what stands out to you as the main components of this pipeline?

Interactive Learning

The AI provides feedback, shows GitHub Actions equivalents, assigns practice exercises, offers hints, and celebrates successes. It adapts its pace, provides contextual help, and maintains an encouraging tone.

The Creation Process

I started by using Microsoft Copilot to craft the perfect prompt, defining the vision and structuring the approach. Then, I used GitHub Copilot Agent Mode to generate the folder structure, initial READMEs, instruction files, and global configuration. I tested the platform by running training sessions, iterating on instruction files, and expanding the curriculum. I continuously improved it by adding detailed examples, troubleshooting scenarios, and assessment checkpoints.

What This Approach Delivered

Learners get a personalised, always-available, interactive, and practical experience that builds confidence. Organisations benefit from scalable, consistent, cost-effective, and self-maintaining training.

The platform successfully covers:

  • Complete Azure DevOps to GitHub Actions migration
  • Advanced topics like matrix builds, environments, and caching
  • Real-world scenarios across multiple technology stacks
  • Troubleshooting and debugging techniques
  • Team collaboration patterns

Implementation Guide: Building Your Own AI Training Platform

What You’ll Need

  • GitHub repository with Copilot access
  • VS Code with GitHub Copilot extension
  • Basic Markdown and folder structure understanding

Steps

  1. Design Your Curriculum: Create a topic-based folder structure (my-training-platform/01-fundamentals/, etc.).
  2. Create Global Instructions: Define the AI tutor’s overall behaviour in .github/copilot-instructions.md.
  3. Build Topic-Specific Instructions: Write detailed instructions for each topic in .github/instructions/.
  4. Test and Iterate: Run training sessions, refine instructions, and expand content.

Best Practices and Lessons Learned

Designing Your Instructions

  • Be specific in your instructions—clearly outline what the AI should do at each step.
  • Include concrete examples to illustrate concepts and guide learners through tasks.
  • Define the AI tutor’s personality and plan out interaction flows to ensure a consistent, engaging experience.

Structuring Your Content

  • Break content into bite-sized, logically ordered topics
  • Define clear triggers and instructions
  • Include reference materials and quick links

User Experience Tips

  • Set clear expectations
  • Provide intuitive navigation
  • Enable flexibility and gather feedback

The Future of Technical Training?

Using GitHub Copilot as an interactive tutor is a big shift in technical training. By mixing AI’s accessibility with solid curriculum design, we create learning experiences that are more engaging, scalable, adaptive, and practical than traditional methods. AI doesn’t replace human instructors; it enhances good instructional design, making high-quality technical training accessible to everyone. The future of technical training is interactive, intelligent, and incredibly engaging.


Try It Yourself

Want to try this out yourself? I’ve put together a fun example training platform that shows off the concept: “Migrating from jQuery to Modern JavaScript”.

This example platform includes:

  • 5 interactive lessons covering DOM manipulation, events, AJAX, animations, and a final challenge
  • Complete AI instruction files that guide learners through each topic
  • Hands-on code examples and practice exercises
  • The nostalgic fun of modernising classic jQuery patterns

You will find the complete source code and instructions in this repository. Feel free to fork it and tweak it for your own training needs—whether it’s framework migrations, adopting new tech, or any other technical learning goal!

The future of technical training is interactive, intelligent, and incredibly engaging. Give it a try!