Inspiration

When firefighters came to our school to give a CPR lesson, the experience left us feeling underprepared and uncertain. We realized that traditional, one-time demonstrations aren’t always enough to make people feel truly confident in a crisis. We wanted to create a tool that allows anyone to practice and learn these life-saving skills at home. By providing real-time, interactive feedback, Briff is built for people who want a private, comfortable way to master CPR and other essential medical checks until it becomes second nature.

What it does

Through using Mediapipe, it first gives instructions to conduct CPR and Pulse checking. Then it can recognize body parts and wait for a person to get into the correct position, where there is an identified victim/person. From there, it will turn green as you get into the correct position to perform the maneuver.

How I built it

We built the core intelligence of the application using MediaPipe for advanced pose estimation and body tracking. To ensure the compression depth and rate were accurate, we integrated a hardware sensor element that communicates directly with our software ecosystem. The entire backend architecture and application hosting are powered by Google Cloud Platform (GCP), creating a seamless bridge between physical hardware data and real-time software feedback.

Challenges I ran into

The mediapipe wouldn't calibrate to the body of a person while moving, which also wouldn't allow it to be placed on the victim's body.

What I learned

We gained hands-on experience deploying and hosting infrastructure on Google Cloud, and we mastered the complexities of syncing real-time hardware telemetry with a software-based computer vision application.

What's next for Briff

We plan to scale Briff to support a much wider array of medical emergency use cases. Our ultimate vision is to deploy this technology across broader healthcare and public safety sectors, equipping everyday individuals with the self-awareness and guidance needed to navigate high-stakes, dangerous situations.

Accomplishments that I'm proud of

  1. Hardware-Software Integration: Successfully mapping a MediaPipe tracking line to a user's body while simultaneously processing hardware data to track CPR speed and depth.
  2. Effective Team Synergy: Balancing effective task delegation based on everyone's individual strengths, which allowed us to move incredibly fast, collaborate smoothly, and make great friends along the way.
Share this project:

Updates