DietPolice

Sticking to your diet is hard when you're the only one who knows that you're about to cheat on it. The whopper isn't worth the humility of your friend roasting you and showing everyone else!

Comment

Inspiration

I wanted to lose weight but I know that I can't stick to my diet if I do not have someone that is keeping watch, but my friends aren't gonna be with me 24/7. So I wanted a way to alert my friends whenever I am somewhere that I do not wanna be.

What it does

Select a friend to keep you in line. Whenever you enter a restaurant that is not good for your health goals, that friend will receive a message effectively snitching on you saying that you are at a restaurant, discouraging you from going to that place!!

How we built it

Apple Watch Sensors - Uses Sensor Logger to stream heart rate and GPS location data HTTP Server - Flask server receives sensor data in real-time Gemini AI Analysis - Google Gemini analyzes GPS coordinates using Maps data via MCP (Model Context Protocol) Location Detection - AI determines if you're at a restaurant with high confidence Twilio SMS - Sends alert message to your accountability buddy: "Gosh is eating fast food right now. Call him!!!"

Challenges we ran into

Fetching my watch data to my phone was difficult, wasn't a variety of tech to help link the data Dealing with phone number errors, we ran into troubles with setting up proper twilio numbers.

Accomplishments that we're proud of

Real time data streaming analysis. Real time coordinate updating. Text messages sent without wifi. Easy to run

What we learned

Claude is goated

What's next for DietPolice

Improve accuracy for restaurant detection and workout detection; we can utilize a database to keep track of previous logs for restaurants, and gyms.

Built With

Share this project:

Updates