The Lecture
Slides: Download the presentation (PDF)
What are we doing today?
A hands-on session that moves faculty from setting up GitHub, VS Code, and Copilot to using AI agents for practical academic work: building a website, converting a syllabus, analyzing data, and building courseware, and then turning those one-off prompts into reusable instructions, skills, and workflows.
Read the official session description
This hands-on session introduces faculty and academic professionals to using AI agents inside GitHub Copilot in Visual Studio Code to streamline everyday academic work. Participants will learn how to sign up for GitHub, install VS Code, activate GitHub Copilot, and access available AI agent features. The session will demonstrate how AI agents can support practical academic workflows, including updating a syllabus, building a simple course website, conducting data analysis, and assisting with automatic grading workflows. Beyond one-time prompting, the session will emphasize how faculty can create reusable instructions, skills, and structured workflows for repeated academic tasks. The goal is to help participants move from casual AI use to more systematic, reliable, and repeatable AI-assisted academic productivity.
What are we going to learn today?
Five things, in this order. The session follows it, and so does this site.
🧠 THE BRAIN Agents · Chat · Models · Tokens · Context
↓
🖥️ THE INFRASTRUCTURE Text Editor · VS Code · Repository · Git · GitHub · Copilot
↓
🎯 THE TASK Goal · Context · Constraints
↓
🔄 THE WORKFLOW Ask · Plan · Agent · Review · Skills
↓
🚀 THE EXAMPLES Website · Syllabus · Data Analysis · Courseware
Parts 1 to 4 are on the Explanation page. Part 5 is the Examples page.
- Part 1: 🧠 The Brain. What's doing the thinking? What an agent is, how it's different from a chat, what a model is, which models are out there, and what tokens and context mean.
- Part 2: 🖥️ The Infrastructure. Where does the agent work? Text editors, VS Code, where your files live, repositories, Git, GitHub, and GitHub Copilot. The setup guide is here too, so you can install everything before or after the session.
- Part 3: 🎯 The Task. How do you tell an agent what you want? You give it a goal, some context, and a few constraints.
- Part 4: 🔄 The Workflow. How do you get a good result more than once? Ask, Plan, Agent, then review. Also
AGENTS.md, saved prompts, skills, and what agents are good and bad at. - Part 5: 🚀 The Examples. Four finished projects you can open: this conference website built from an outline, a Word syllabus turned into HTML, one dataset analyzed by two different AIs, and interactive courseware.
Why am I doing this?
Because this changed things for me.
I always wanted to build things but didn't feel I had the knowledge. Without being an engineer, I thought I couldn't do it. Then AI agents changed everything. Now I build websites, scripts, automations, you name it, and I'm happy to share that knowledge.
This website is the presentation, and AI agents helped me build it. So I'd rather show you than tell you.
Why it matters
You probably use AI already. You open a chat, ask something, copy the answer, and paste it somewhere. Next week you do it all over again.
That's fine for a quick question. It's not great for the work that keeps coming back: the syllabus you update every semester, the course site, the data you look at after every term.
Today is about setting that kind of work up once. You check the result, and next semester you run it again instead of starting over.
You don't need to be a programmer for any of this. If you can describe what you want and check what you got, you can do it.
How to use this site during the session
Follow along on your phone.
Every section has its own link, so you can jump to whatever we're talking about.
The Examples page has the prompts and files I used, so you can try everything later.
Next → Part 1: The Brain