Overview
Gooblet came out of the Winthrop Vibe Hackathon as our attempt to rethink personal finance for college students who avoid traditional budgeting apps. Most finance tools bury people under dashboards, charts, and categories they never want to look at. We wanted something faster, more opinionated, and a lot more human.
Instead of asking users to manage their entire financial life in-app, Gooblet centered everything around one question: how much is safe to spend today? That became the product. One number, delivered clearly, backed by live financial data and wrapped in a personality-driven interface that made the signal feel alive instead of clinical.
The Core Product Idea
The strongest insight behind the build was that nobody actually opens finance apps regularly unless something has already gone wrong. That meant building a better app alone was not enough. The information needed to come to the user instead of waiting for the user to come check it.
That is why we built a home screen widget through Scriptable. It surfaced the safe-to-spend number and Gooblet's current mood directly on the user's iPhone home screen, with no app launch required. The result was a finance experience designed around passive visibility instead of active maintenance.
What We Shipped
In 24 hours, we got a surprising amount working:
- A safe-to-spend calculator with danger-day predictions
- A 3D procedural mascot built with Three.js
- Live bank integration through Plaid
- Full authentication, transaction syncing, and risk classification
- An iOS home screen widget through Scriptable
- A React, TypeScript, and Tailwind frontend with a Node.js, Express, and Prisma backend
The character was not just visual flavor. Gooblet's emotional state changed based on the user's financial habits, which gave the interface a lightweight feedback loop and made the app feel memorable in a way most finance tools do not.
My Role
I served as team lead and backend developer. My main ownership areas were the API architecture, Plaid integration, authentication system, and database layer. A lot of the challenge was not just writing the backend itself, but making sure the data contracts were good enough for the frontend, the widget, and the financial logic to all move together under a very short deadline.
The team moved fast, and I had a lot of help from them along the way, but my focus was on making the backend reliable enough that the rest of the experience could exist on top of it.
Why It Matters
Gooblet is one of my favorite hackathon builds because it was not just a pile of features. It had a real product insight, a clear user behavior assumption, and a technical implementation that actually matched the idea. It is the kind of project I like most: fast-moving, a little weird, technically demanding, and grounded in a problem people actually have.
