Teams using AI heavily are seeing something: developers accept code without understanding it, juniors skip learning steps, seniors stop writing code themselves. Skills that aren’t practiced fade – quietly, slowly. By the time you notice, it may be hard to fix.
But teams that avoid AI fall behind. So what do you do – miss the benefits, or risk digging a hole you can’t climb out of?
This seminar introduces Controlled AI-Assisted Development – a practical way to use AI tools while keeping code quality high and helping your team grow. You’ll learn all of GitHub Copilot‘s customization features: instructions, slash commands, custom agents, skills, MCP tools, and hooks. For each one, you’ll see how it works, when to use it, and when to skip it.
We use VS Code and GitHub Copilot for demos, but the ideas work for Claude Code, Cursor, and other AI coding tools too.
This isn’t just a lecture. In the discussion parts, we’ll figure out the opportunities and risks together – mixing the instructor’s experience with your real questions. The goal is to leave with a practical approach that works for real teams.
