Build inside the gap,
not above it.
The best cardiac monitoring sits in hospitals — or on $800 watches. The people most likely to collapse in public can't reach either. SamaBeat is built for everyone the wearable industry priced out and miscalibrated for.
Four years inside the industry
that priced this out.
Four years inside cardiac and neurostimulation at Boston Scientific, Medtronic, and Abbott surfaced the same pattern: the most advanced monitoring is locked behind hospital walls or an $800 price floor, and the populations with the highest cardiac burden are systematically underrepresented in the datasets these devices were calibrated on.
SamaBeat is the product we would have wanted available to our own families — bystander-readable in the 10-second decision window, accurate on the skin tones the industry has historically ignored, and priced for the people the $300 floor has structurally excluded.
Calibration-agnostic posture + heart-failure prediction. Two patent approvals.
ML extraction of nerve signal from NIM Vital. Cross-functional complaints interface.
C++ on the CardioMEMS heart-failure monitoring device. Signal characterisation + circuit-board rework.
Physics-informed deep-learning for low-dose CT. Gait analysis under Dr. Gari Clifford. Diagnostics 2020.
Two builders.
One constraint.
Aranyo converts research to repeatable pipelines. He previously built Autest, a culturally adapted autism screening app that placed 1st with the American Psychological Association, and shipped AI products at Graymatics, notably impacting 700k+ daily riders on the Bangalore Metro. His work has garnered $180k+ in revenue across 5 startups. Beyond code, Ray created and scaled an US Air Force-commended nanopesticide to 1800+ farmers, and received $14k+ in funding as a Wu Tsai Scholar at Yale. At SamaWritten, Ray leads partnerships, clinical trial pathways, and product. He has helped scale his family co-owned Anubhav, an NABL-accredited diagnostic in India's most populous district, to $1M ARR in 3.5 years. Automating operations and improving efficiency with data led him to found a low-cost, clinical-grade wearable for disproportionately burdened populations. Ray holds a BS in Computer Science and a BS in Psychology from Yale and is excited to lead a revolutionary new foray in cardiac care, with tech as living as the person it helps.
- ✓First Place, APA ISEF 2021
- ✓700K+ metro rider deployment
Pushti builds ML algorithms and signal processing pipelines for implantable and diagnostic medical devices. At Boston Scientific, she has received approvals for 2 patents and 20+ invention disclosures, and been recognized with a Top Oral Presentation award and Featured Inventor Program. With a focus on heart failure, her collaborations span heart rhythm management, neuromodulation, urology, and DENALI. She previously developed ML tools at Abbott for CardioMEMS and Medtronic for the NIM Vital nerve monitoring system, among others. At SamaWritten, Pushti writes code for signal processing, person-level calibration, clinical validation and the web and mobile apps. Her expertise lies in using existing datasets to debias and finetune algorithms; for example, she has repurposed cardiac sensors to track menstrual cycles, menopause, and cyclical AF risk spikes, with no precedent or funding. Pushti holds an MEng (Duke) and BE (GTech) in Biomedical Engineering, with a Certificate in Biomedical Data Science. After a cardiac physiology immersion at UMinnesota in May 2025, she dived deep into medical bias in health monitoring, specifically in underrepresented groups, and can't wait to impact millions with algorithms that understand them.
- ✓BS, Georgia Tech (3.87)
- ✓MS Duke + Cert. Biomedical Data Science (3.85)
Three rules.
Every choice falls out of them.
Start with the constraint, not the solution.
The hardest constraint in cardiac monitoring isn't sensor accuracy. It's the $300 price floor that excludes the bottom 70% of the at-risk population — by income, geography, and language. Every architectural choice in SamaBeat — e-ink display, sub-$10 BOM, on-device inference — falls out of refusing to design above that line.
Build inside the gap, not above it.
The populations with the highest cardiac burden — South Asian, Black, low-income elderly living alone — are the same populations underrepresented in the calibration datasets every consumer wearable depends on. The FDA confirmed in 2022 that pulse oximeters overestimate SpO₂ on darker skin. We chose to validate on Monk Skin Tone 4–10 first, not last.
A tool used daily beats a tool that demos well.
Continuous wear, bystander-readable, no app, no passcode, no battery anxiety. The same sensor stack that identifies the patient to a stranger captures the cardiac signal that predicts the collapse making the ID necessary. One device, two clinical jobs, one price the structurally excluded can pay.
Three curves
finally crossed.
None of these technologies are new. The convergence is. Five years ago this device could not have shipped at this price. Five years from now the calibration data we are gathering today will be the moat.
Always-on, zero battery cost
Reflective displays mature enough to hold bystander-readable text indefinitely. The only display class that closes the 10-second decision window with no app, no button, no unlock.
Clinical sensors, consumer cost curve
The same sensor class that lived only in $800 wearables now ships at sub-$10 BOM. The price floor for continuous, accurate cardiac monitoring collapsed under a sub-$70 retail.
Arrhythmia detection, on-device
Cardiac event prediction no longer needs a cloud round-trip or a clinician in the loop. Low-power MCUs run the models locally — the same transition the founder built inside at Boston Scientific.
The technical risk
is retired.
Cross-sectional clinical validation completed Oct–Dec 2025 at an NABL-accredited diagnostic facility, deliberately recruited across Monk Skin Tone categories four through ten — the exact range underrepresented in every major calibration dataset in consumer wearables.
Hardware works.
Validation works.
Distribution is next.
- ✓4 years across Boston Scientific, Medtronic, Abbott — DCRM, NIM Vital, CardioMEMS
- ✓Tonsiliscope: non-provisional patent filed, moving toward commercial release
- ✓182-participant validation closed; observational protocol SWHT-SAMA-001-2026 filed
- ✓Stanford Cardiology international co-investigator confirmed
- ✓Dual IRB tracks live: Mumbai (CritiCare Asia, 1,200+ beds) and Indiana (US)
- ✓Sub-$10 BOM, sub-$70 retail target, 14+ days battery, zero cloud dependency
The data is the moat.
And it compounds with every wrist.
Every SamaBeat worn becomes an anonymised data point in the first cardiac dataset built deliberately on underrepresented skin tones. That dataset is how the calibration problem gets solved at the industry level — not only for our device, but for every pulse oximeter that ships after.
One product. A low-cost consumer device, a public-health instrument, and a research dataset, in one piece of hardware. It scales the way phones did — cheap enough to reach the next billion users, not only the first.