LyraCare

Inspiration

Millions in India lack access to early detection tools for serious health events. Seeing this gap motivated me to create a solution that leverages widely available smartwatches to provide real-time health monitoring and emergency alerts.

What it does

LyraCare connects to a user’s smartwatch and continuously monitors vital signs like heart rate, SpO₂, and ECG. Using a machine learning model trained on medical datasets, it detects emergency-level anomalies and sends alerts to the user and their emergency contacts to reduce response times.

How we built it

I built the mobile app with React Native and integrated smartwatch APIs for real-time data collection. The ML model was developed using Python and TensorFlow, trained on medical data. For deployment, I used Azure Cloud with Kubernetes for scalability.

Challenges we ran into

Balancing accurate anomaly detection with minimizing false alarms was difficult. Integrating diverse smartwatch sensor data and ensuring timely notifications also required extensive testing and optimization.

Accomplishments that we're proud of

Successfully developing a prototype that connects to smartwatches and performs real-time anomaly detection, providing timely alerts that could potentially save lives.

What we learned

I gained hands-on experience in wearable data integration, real-time ML inference, and deploying cloud-based health applications. The project deepened my understanding of healthcare technology challenges.

What's next for LyraCare

Improving model accuracy with larger datasets, expanding compatibility to more smartwatch brands, and developing additional features like predictive health insights and user-friendly alert customization.

Built With

• amazon-dynamodb
• amazon-sns
• amazon-web-services
• aws-iam
• aws-lambda
• kubernetes
• python
• react-native
• tensorflow