HuskyHack 2025

Nudge

Smart Savings for Sound Credit Union

A mobile banking feature for Sound Credit Union that surfaces personalized deals, tracks savings, and delivers AI-powered financial advice — all inside the app members already use. Built at HuskyHack 2025 for the Sound Credit Union fintech challenge.

Timeline Nov 2025
Role Full-Stack Developer
Team Hackathon — HuskyHack

The Challenge

Credit union members leave money on the table every day. Cashback offers go unnoticed, local deals expire unclaimed, and loyalty points pile up unredeemed. Banks have the transaction data to act on this — but most don't. The challenge was to turn existing spending history into a proactive savings engine without requiring members to change apps or habits.

The Solution

Nudge lives inside the Sound Credit Union app as a dedicated tab. It analyzes spending history to surface relevant deals, fires real-time location-based notifications when members are near partner merchants, and generates personalized savings insights using GPT-4o-mini — turning passive transaction data into active financial guidance.

Key Features

Personalized Deals Marketplace

Members see a curated feed of offers matched to their spending habits — local coffee shops, grocery cashback, travel rewards, loan promotions, and limited-time bonuses. Deals are filterable by category and sorted by relevance. Tap any deal to see full details and a QR code for instant in-store redemption.

  • Spending-history-matched offer feed
  • Category filtering (Food, Travel, Loans, Shopping)
  • QR code for instant redemption

Location-Aware Nudges

When a member is near a partner merchant, Nudge fires a real-time in-app notification — "You're 0.4 mi from Stumptown Coffee — get 25% off now." No separate app, no opt-in friction. The right offer appears at exactly the right moment.

  • Proximity-triggered in-app alerts
  • Partner merchant network integration
  • Zero extra app installs needed

Savings Dashboard & AI Insights

A dedicated dashboard shows total saved this month with month-over-month trend, a 5-month savings history chart, and a category breakdown. GPT-4o-mini generates personalized tips from the member's actual savings profile, plus a projection of what current habits are worth over time.

  • 5-month savings history with Recharts
  • GPT-4o-mini personalized advice
  • Forward-looking savings projection

Technology Stack

Frontend

React 18 TypeScript Vite Tailwind CSS

UI Components

shadcn/ui Radix UI Recharts Motion

Backend & AI

Node.js Express OpenAI GPT-4o-mini

Deployment

Docker + Nginx GitHub Pages

Technical Highlights

3
Core pillars: personalized deals, location nudges, and AI savings insights
4
Deal categories — Food, Travel, Loans, Shopping — each with filterable offer feeds
1
AI endpoint: GET /api/savings/analysis returns GPT-4o-mini insight from live savings profile
5
Months of savings history charted with trend comparison and forward projection

Key Learnings

01

LLM for Personalization at Scale

GPT-4o-mini was ideal here — cheap enough to call per user session, fast enough to feel real-time, and capable enough to generate genuinely useful financial advice from structured savings data.

02

Location Context Changes Everything

A deal shown when someone is 0.4 miles away is fundamentally different from the same deal shown at home. Proximity-based nudges dramatically increase the perceived relevance and actionability of offers.

03

Mobile Phone Mockup UX

Rendering the app inside an iPhone frame in the browser made demos instantly intuitive for judges and stakeholders — no installation needed, and the form factor communicated "mobile app" immediately.

04

TypeScript for Rapid Hackathon Dev

Strict TypeScript caught several subtle data-shape bugs during the hackathon that would have been hard to debug under time pressure. The upfront investment in types paid off in fewer runtime surprises.

Interested in Collaboration?

I'm always open to discussing fintech projects, internship opportunities, or AI-powered product ideas.