Where it came from
I used to teach, and I've spent a long time on how curricula are designed: what someone needs to understand, in what order, and how you know they've got it. I wanted to learn IT, properly, across the disciplines people actually work in, and I didn't know how to go about it. Courses taught the tools; I wanted to learn the work.
So I combined the two. I took the way I knew how to build a curriculum and used AI to build the one I wanted to take: real projects for realistic clients, where you direct the AI through the work and check what it produces, which is how the work is increasingly done. It grew into seven tracks and 165 projects.
Who did what
I set the direction: what the platform is for, the tracks, the teaching framework, and the rules every project has to follow. I reviewed the design of each track and project, and made the calls on what is taught.
Claude, Anthropic's AI model, did the rest: researched each discipline, designed the curricula, wrote every project and its materials, built this site, and runs the checks that keep it working. A second model, OpenAI's Codex, reviews the finished work independently. The client and the senior colleague you can talk to inside a project are AI too, and the product photographs and portraits are generated images.
The clients are fictional
Every project is for a fictional client in a real industry, grounded in how such businesses actually work. No real company's data, documents or staff appear anywhere.
Independence
This platform is independent. It adapts the Teaching for Understanding framework developed at Harvard's Project Zero and has no connection to Harvard or Project Zero. It is built with models from Anthropic and OpenAI and has no connection to either company. None of them has seen it, reviewed it, or endorsed it.
It is free, with no paid tier and no advertising. If something doesn't work, the feedback button on any page reaches a person.
Contact: josh@torkington.au