Inspiration
Heart failure remains one of the leading causes of hospital readmission worldwide, resulting in $284 billion annually spent on heart failure treatement. Monitoring is often reactive and complications are detected only after symptoms worsen. We were inspired to build a system that shifts care from hospital-based reaction to continuous, predictive prevention all on your wrist.
What it does
Ventria is an AI-powered wearable platform that continuously monitors heart failure patients and individuals on high-risk cardiac medications. It tracks physiologic signals 24/7 and integrates medication, lab, and symptom data to generate real-time risk scores for decompensation, toxicity, arrhythmias, and renal-electrolyte complications. This allows the patients to get full reports within their dashboards for their heart vitals.
How we built it
We designed a wearable-integrated digital platform consisting of continuous vital tracking (heart rate, HRV, rhythm patterns, activity, sleep), patient-entered medication and lab data, and a cloud-based AI risk engine. We developed logic models that analyze deviations from individualized baselines and medication-specific risk factors to generate explainable, tiered alerts for patients and clinicians.
Challenges we ran into
Defining clinically meaningful thresholds without creating alarm fatigue Balancing medical depth with a simple, usable interface Integrating medication safety logic with physiologic trend monitoring Designing a system that works for both high-risk patients and general users Figuring a way to integrate the hardware with the software side
Accomplishments that we're proud of
Built a concept that combines wearable physiology with medication safety — not just fitness tracking Designed explainable AI results rather than black-box alerts Created a scalable solution targeting both clinical and consumer markets Developed a product vision focused on prevention, not reaction Figured a way to integrate certain sensors within the band for the heart tracking listed Understood the methods needed to integrate completely with iOS Apple Watches
What we learned
We learned that heart physiology is far more predictive than we initially realized. Subtle changes in heart rate trends, heart rate variability, respiratory rate, activity levels, and sleep patterns can signal worsening heart failure or medication toxicity days before hospitalization. We were surprised that despite the availability of wearable data, there is still no widely adopted system that integrates continuous vitals with medication profiles and lab values to predict risk in real time. We also gained a deeper appreciation for how powerful heart rate dynamics are not just as a fitness metric, but as an early warning of systemic instability. The opportunity to use everyday wearable data for clinical prevention is largely untapped.
What's next for Next,
we aim to validate Ventria in collaboration with cardiology and heart failure specialists, integrate real wearable device APIs, and refine our AI risk models using real-world clinical datasets. We plan to explore regulatory pathways for medical-grade monitoring, pilot the platform within remote patient monitoring programs, and further develop clinician dashboard capabilities to support proactive, value-based cardiac care. Within the next 2 weeks we want to finish working on the hardware prototype of the band to create even more accurate data.
Built With
- claude
- javascript
- loveable
- python
Log in or sign up for Devpost to join the conversation.