CardGenie

Inspiration

We noticed something frustrating - Our friends have multiple credit cards but still use the wrong one half the time. One has a 5% grocery card but uses their 1% card at Safeway. Another let 60,000 airline miles expire. Americans collectively leave $30 billion in rewards unclaimed every year not because they don't care, but because tracking it all is exhausting. We thought: what if your phone just told you which card to use, right when you needed it?

What it does

CardGenie is your AI co-pilot for credit card rewards. It analyzes all your cards, tracks your loyalty programs, and tells you the optimal card to use at every purchase. Walk into Target? Get a notification: "Use Chase Freedom here 5% back vs 1%." It also prevents value loss with expiration alerts, shows you exactly how much money you've been leaving on the table, and gives personalized tips like "activate your Q4 bonus" or "pay down this card to boost your credit score."

How we built it

Started by deeply understanding the problem, researching users pain points, studying redemption rates, and mapping out where people lose money. Then designed an architecture that will connect via Plaid API to pull transactions, use a rules-based recommendation engine (with plans for ML later), and integrate with AwardWallet for loyalty tracking. The tech stack is React for WebApp, Node.js backend, PostgreSQL database, and GPT-4 for the AI chat assistant. Focused Phase 1 on the architecture, user flows, and proving the concept could work for Sound CU members.

Challenges we ran into

The biggest challenge was complexity. Credit card rewards rules are a maze, rotating categories, spending caps, partner transfers, expiration dates. Building a system that accounts for all these variables while keeping the user experience dead simple was tough. Also had to figure out real-time merchant detection without draining phone batteries. And honestly? Prioritizing features. Had 20 ideas but only time to build the essentials first.

Accomplishments that we're proud of

I am proud that I didn't just ideate a trackee but a complete optimization platform. The competitive analysis showed us that nobody combines tracking + real-time recommendations + expiration alerts + credit monitoring in one place. Also proud of the user personas created. they're based on real stories (Source: Reddit), and the solution that could genuinely solves their problems. Plus, I designed it to be an ease to use and adopt app, making it accessible to all members from day one.

What we learned

I learned that solving a common problem isn't enough, you have to solve it in a way that's easier than the current workaround. People already track rewards manually in spreadsheets. CardGenie has to be 10x better to win. Also learned that financial APIs like Plaid make this possible in ways that weren't feasible five years ago. And most importantly: people want to optimize their rewards, they just need someone to do the thinking for them.

What's next for CardGenie

First, we win this hackathon and partner with Sound CU for a beta launch. Phase 2: build the MVP with core features. Phase 3: launch to 1,000 members and validate product-market fit. Phase 4: scale to multiple credit unions and add machine learning to make recommendations even smarter. Long-term vision? CardGenie becomes the standard tool every credit union offers their members, turning rewards confusion into automatic value capture. Want to help Americans reclaim that $30 billion, one notification at a time.Retry

Built With

• claude
• css
• lovable
• react
• sora
• typescript