DoubleTake

Inspiration

DoubleTake started with a worry many families share: older relatives are often targeted by financial scams, but it can be hard to know when something is wrong or how to bring it up without making them feel judged. We wanted to create a tool that helps families notice unusual activity and talk about it with care, while keeping the account holder in control.

What it does

The app uses simulated transactions from Capital One’s Nessie sandbox to flag activity such as unusual withdrawals, large transfers, and gift card purchases. Each alert explains what caught the system’s attention. A relative can check in, and the account holder can set their own preferences. We also built a community page for scam reports and a trustee view for reviewing activity.

How we built it

We built DoubleTake with React and TypeScript on the front end and Express on the back end. Our alerts come from rules we wrote, not an AI service. Nessie gave us realistic sandbox data to work with, without connecting to real bank accounts or collecting real banking credentials.

Challenges we ran into

One challenge was making alerts useful without making them frightening or accusatory. We learned that people need to see the reason behind a warning and have a comfortable way to respond. We also wanted to be honest about what we built. DoubleTake is a prototype, and its bank connection is only a demo. It does not access real accounts or move money.

What's next

DoubleTake is a prototype, and there’s plenty we’d like to improve. We’d like to explore secure bank connections, with clear consent and privacy controls, so families can use real account activity instead of sandbox data. We’d also like to make alerts more customizable, improve accessibility for older adults, and support secure sharing between family members on different devices. We’d want to test these changes with older adults and their families to make sure the app feels useful, respectful, and easy to understand.