Inspiration

Cardiac arrest is one of the most time-sensitive medical emergencies, yet bystanders often don’t know what to do or don’t get alerted in time. We were inspired by one core idea: if CPR starts in the first few minutes, survival chances rise dramatically. HandsForHearts was built to close that gap by turning nearby people and devices into an immediate response network.

What it does

HandsForHearts is a rapid-response cardiac emergency platform. It combines wearable-based emergency detection, nearby responder alerting, and guided CPR support. In our prototype flow, emergency signals can trigger alerts, notify nearby responders, provide location context, and walk responders through clear CPR steps so help starts before EMS arrives.

How we built it

We built HandsForHearts as a multi-part system:

  • A Next.js + Tailwind web experience to communicate the mission and simulate the response flow.
  • A Swift watchOS + iOS MVP using HealthKit, CoreMotion, and WatchConnectivity for detection handoff, cancellation window logic, and emergency relay behavior.
  • A FastAPI backend for ECG image analysis and alert routing, with WebSocket/polling-based delivery for iPhone clients.
  • A React Native iOS client path that receives emergency events and triggers local notifications.
  • An ElevenLabs emergency relay module to generate provider-ready emergency handoff messaging.

Challenges we ran into

Our biggest challenge was reliability in real-world mobile conditions—especially emergency alert delivery when apps are backgrounded or closed. We also had to balance detection sensitivity with false-positive risk, while handling complex permission/entitlement requirements (health data, location, critical alerts). Finally, integrating multiple platforms in hackathon time pushed us to design strong fallbacks and modular components.

Accomplishments that we're proud of

We’re proud that we built more than a concept—we built a working prototype across frontend, mobile, and backend layers. We implemented an end-to-end emergency flow with detection handoff, responder-facing alerts, and CPR guidance. We also shipped practical fallback communication paths and a provider-relay component, which makes the system feel actionable in real emergency scenarios.

What we learned

We learned that emergency tech is as much about trust and UX clarity as it is about AI and sensors. Every second, tap, and message matters under stress. We also learned the importance of designing layered reliability (multiple delivery paths) and using a phased architecture so we can iterate safely from prototype to real deployment.

What's next for HandsForHearts

Next, we want to harden HandsForHearts into a production-grade platform: deploy robust geofenced alerting, implement APNs Critical Alerts end-to-end, and add live location streaming at scale. We also plan to expand responder tooling with AED guidance, multilingual CPR instructions, and stronger emergency-service integration. Long-term, we want to validate performance with pilot users and bring the system to both iOS/watchOS and Android ecosystems.

Share this project:

Updates