Inspiration
We saw that backend security alerts fail when scammers hack human psychology, pushing members to willingly send their own money out of panic. Because scammers adapt faster than code, we wanted to build "psychological immunity" by letting users rehearse the scam before it happens
What it does
Otterly Safe is a gamified SDK that plugs directly into existing credit union mobile banking apps. It uses an animated otter coach named Ollie to guide users through mini-games to spot red flags, interactive role-plays with fake scammers, and quick quizzes that reinforce lessons on urgency and secrecy
How we built it
We structured it as a white-label B2B plug-in SDK designed for credit unions. The mobile app integrates with the Otterly Cloud to pull a library of synthetic scam scenarios, while aggregated completion and quiz results are sent to a dashboard for the credit union to monitor.
Challenges we ran into
A major focus was balancing realistic scam simulations with strict user privacy. We had to design an effective intervention system that relies entirely on synthetic, pre-scripted scenarios without ever reading a user's texts, listening to their phone calls, or tracking their device activity
Accomplishments that we're proud of
We successfully developed a clickable prototype featuring Ollie and our first set of simple scam scenarios. We are also proud of structuring the business value to launch directly with Sound Credit Union's 184,000+ members, while remaining scalable for over 4,200 federally insured credit unions.
What we learned
We learned that interactive practice significantly outperforms reading static material. While annual security PDFs and plain text warnings are often ignored, interactive interventions can cut scam completion rates from 22% down to just 4%.
What's next for Otterly Safe
Over the next one to four months, we will expand our library with more mini-games, role-plays, and quizzes covering newly emerging scam types. Between months five and nine, we plan to implement rich Ollie animations, a rewards and streaks system, Spanish language support, and AI-drafted scenarios
Built With
- framer
- next.js
- openai
- postgresql
- python
- react
Log in or sign up for Devpost to join the conversation.