BUILT WITH:
ChatGPT
Codex
React + TypeScript
GitHub
Vercel
Figma
THE EXPERIMENT
I wanted to understand what changes when a product designer moves beyond prototypes and starts working directly in code.
I designed and built a mobile-first budgeting app around a simple idea: instead of tracking what you've already spent, tell you what you can safely spend today.
Using AI-assisted development, I took the product from concept and interaction design through React implementation, Git-based development, QA, and a live production deployment.
Your monthly discretionary budget becomes a daily spending target, automatically adjusting as you spend, save, and move through the month.
The experiment became a fully functional client-side web app with real calendar logic, persistent browser storage, transaction history, editable past days, monthly/category summaries, responsive states, and production deployment.
Learnings:
Building in code changed the level at which I could evaluate my own designs. Decisions about state, persistence, responsive behavior, and edge cases stopped being handoff details and became part of the design itself.
More importantly, it gave me a better understanding of how to use AI as a bridge between design and engineering — not to replace either discipline, but to shorten the distance between an idea and something people can actually use.
Up next: the MVP is live.
Now I’m exploring what it takes to turn it into a real product.
Onboarding + Landing Page
Design a lightweight experience that explains the budgeting model and gets users to their first daily number quickly, alongside a landing page (featured above) to invite users to try it out.
Backend + Accounts
Move beyond browser-local storage with authentication, persistent user data, and syncing across devices.
Native App
Explore translating the responsive web experience into a native iOS product and taking it through the process of launching on the App Store.
Keep Learning
Continue using the project as a testing ground for AI-assisted product development, moving further into implementation while staying user-focused.