Sin City Saviors

Inspiration

Growing up in Las Vegas, we have been surrounded by gambling addiction in our friends and families. For this hackathon, we decided to build an impactful project that can measurably track and improve user’s gambling habits, and make steps towards helping their gambling addiction.

What it does

Don’t have a real companion?? Ace will actually be your companion on the go. Ace will monitor win/loss rates, our budget of the day, and as we are playing Ace will always be checking on you. If he sees that you are struggling, he will definitely be a friend and tell you to take a break and try a different scenery.

How we built it

There are three layers working together. First, the onboarding check-in collects mood, stress level, and intent before the session starts. These feed a mental health modifier that adjusts the risk score baseline someone playing to escape stress gets a closer watch than someone celebrating a birthday. Second, the risk engine runs on every bet, scoring four behavioral signals: loss chasing (bet size increasing 1.5x+ after consecutive losses), session duration over 90 minutes, betting velocity doubling in the last 5 minutes, and bankroll erosion of 50%+ within 30 minutes. Each signal carries a weighted score that adds up to a 0 to 100 risk rating. Third, when the score crosses a threshold, Ace, a Claude-powered AI companion automatically sends a personalized message. It's not a pop-up or a warning. It's a casino host suggesting the steakhouse has a table open, or that tonight's show has good seats left. Players can accept the suggestion or dismiss it, and the whole interaction is logged to their session history.

Challenges we ran into

Our biggest challenges were working as a team and deciding on an idea that we were all passionate about. Additionally, it was a struggle because we had a variety of skill levels, so this entire hackathon was a learning experience for us.

Accomplishments that we're proud of

We’re proud of our risk management algorithm and authentication with real emails, as well as our several simulation of different kinds of behaviors in gambling.

What we learned

We learned that gambling can be a huge problem and while we don’t want to roughly enforce all gamblers to quit, we want to make sure that gamblers are able to safely monitor their gambling budget and always have a mental check in.

What's next for Sin City Saviors

The one honest limitation for us is that this is a simulation. For the future, real world accuracy would require calibration against actual player data, which is the next step for any casino operator deploying this in production.

Built With

• anthropic-claude-api
• github-oauth
• google-oauth
• neon-(postgresql)
• next.js-16
• nextauth.js-v5
• prisma-7
• react-19
• tailwind-css-v4
• typescript
• vercel
• vercel-ai-sdk