There's no throwaway practice project. The app you start in week one is the app you demo to employers in week twelve — only twelve times more capable. Each new skill enters exactly when your app needs it.
| Week | Your app gains… | …so you learn |
|---|---|---|
| 01 | A home on the internet | Git & deployment |
| 02 | Memory — data that persists | databases |
| 03 | Users & privacy | authentication & security |
| 04 | Speed & reliability | caching, performance, error handling |
| 05 | A brain | building on LLM APIs |
| 06 | Knowledge of its own data | RAG — retrieval over your data |
| 07 | The ability to act | AI agents & tools |
| 08 | Proof that it works | evaluations & rigor |
| 09–12 | Production polish & a public launch | scope, hardening, demo day |
Take an idea from a blank screen to a real, multi-user app on the live internet. By the end of this arc, shipping a full-stack feature is second nature.
Add genuine AI features that are reliable, not just demos — a feature that understands your data, takes action, and that you can prove actually works.
Scope it, harden it, polish it, and launch — then present it on Demo Day to your cohort and invited employers. You own it end to end.
Every Friday ends with something live and a demo you can defend.
Pick the week's build and list the unknowns you'll have to figure out.
Build it into your app with an AI coding agent. Keep a learning & debugging log.
Deploy. 10-minute demo to your cohort. Mentor review + “explain it back.”
Small cohort, mentor-led, built entirely on free-tier tools.
Apply for the next cohort → See the tools & differentiation