48 Hours, 100+ Teams, One P2P Lending App: Winning Chain Fusion with Creditopia

The brief that filtered out most teams

The prompt was not "build something cool with ICP." It was: solve a real traditional-finance problem, and deliver impact to end users — not just impress the judges. That framing quietly eliminates half the field, because it rewards teams who can name a real user and punishes teams who lead with technology.

We picked SME financing. Small businesses that banks find too small or too risky, and lenders who would fund them if they could reach them safely. Creditopia is the bridge: peer-to-peer lending on the Internet Computer, where the smart contracts handle the financing and every transaction lives on-chain.

Why ICP, and what it cost us

Building on the Internet Computer meant Motoko and Rust for the canisters, with JavaScript and Vue.js on the front. The upside is that a lot of the app can live fully on-chain. The cost is that the toolchain is less traveled than EVM — fewer StackOverflow answers at 3 a.m. That trade is fine when the whole point is to show what the platform can do.

The decisions that actually won it

The code mattered less than three decisions we made under pressure:

1. 1Scope down, hard. We shipped the lending loop end to end rather than five half-features. A judge can forgive a missing feature; they cannot forgive a demo that breaks.
2. 2Protect the demo path. By hour 30 we stopped adding and started hardening the exact flow we would show on stage.
3. 3Manage energy like a resource. Forty-eight hours is a stamina problem disguised as a coding problem. We rotated who slept so the team was never fully down.

The most memorable part was competing alongside teammates from completely different backgrounds and campuses, making big decisions under real time pressure. The win was almost a side effect of getting those decisions right.

What a hackathon teaches that a job cannot

Compressing a real product into 48 hours forces a clarity you rarely get otherwise: what is the one loop that has to work, and what can you cut without lying about what you built? I take that instinct into every client project now. Most software does not need more features. It needs the core loop to be undeniable.

Creditopia finished first out of 100+ teams and raised a $3,000 developer grant. But the transferable win was the discipline — and I have leaned on it in every build since.