Project labs AI Builder Lab
From an idea to a working product
Students don’t just learn to write prompts. They come to understand the path from an idea to a working digital product. AI is not an authority here: it builds something, but students have to understand what happened, verify it, fix it and explain it.
Small project group · up to 12 students
What students do
- They break an idea down into steps and work with prompts, tokens and context.
- They use Git and GitHub, the CLI and APIs, and learn to manage API keys securely.
- They store data in a database (e.g. Supabase) and try out both cloud and local models.
- They deploy the app and explain what AI built in it and what they verified and fixed themselves.
Possible outcome
A working, deployed app or tool, plus a short defence of their own solution.
Subject links
- computer science
- mathematics
- language and literature
- entrepreneurship
Suitable formats
- Workshop 90–120 min (suitable)
- Project day 1 day (suitable)
- Intensive programme 2–5 days (suitable)
- Project block weeks (less suitable)
- External studio semester + (suitable)
- Online / hybrid possible
Technical requirements
A computer classroom with internet access. We sort out accounts and API keys together, in line with the school’s rules and the students’ age.
Group size
Small project group: ideally 6–10 students, 12 at most.
A smaller group leaves room for individual guidance, conversation, experimentation and genuine teamwork instead of a lecture-style workshop. Larger groups can be worked out case by case, for example by splitting students into several project teams or blocks.