Solutions Deployment Clinical Approach FAQ Company Book a demo

Turn signals into vitals.

A face. A voice. That's all it takes. HeartBit turns a face into vitals, VoiceBit turns a voice into biomarkers, both contactless and real-time, built to integrate into the care platform you already run.

2Signal products, one integrated platform
<30sTypical capture time, from a single short recording
0Wearables required, or extra setup ever required
Real-timeSignal inference running on the edge or in the cloud

Solutions

The body speaks. We help you listen.

A face carries a pulse, a voice carries a breath. HeartBit and VoiceBit read both, and turn what's already there into a clinical picture. Pick the one that fits your workflow.

HeartBit

Beats per minute, bits per second.

HeartBit turns any camera into a vital-signs sensor, using remote photoplethysmography (rPPG) to read the same subtle skin-color shifts a pulse oximeter detects. No wearable. No contact. Just a face, and a few seconds.

Heart rate Respiratory rate Blood pressure (est.) SpO₂ (est.) Heart rate variability
Explore HeartBit
HeartBit mobile app screen showing a live rPPG capture on a face: heart rate 91 BPM, estimated blood pressure 118/74 mmHg, SpO2 trend 99%, respiration 10.3 per minute, and the underlying rPPG signal chart with Y, G, and POS traces plus an FFT heart-rate histogram.

No wearables · No finger clips · No contact at all. Just look at the camera for under a minute.

What HeartBit measures

Heart rate

Beat-by-beat pulse detection with multi-region signal fusion for stability.

Blood pressure trend

A relative estimate derived from the shape of the pulse waveform.

SpO₂ trend

Blood-oxygen trend estimation from subtle color-channel ratios.

Respiratory rate

Breathing rate inferred from subtle shifts in the pulse signal.

Heart rate variability

A wellness indicator drawn from beat-to-beat timing.

Rhythm screening

A statistical check for irregular heartbeat patterns.

Built for clinical trust

Every reading is cross-checked across six face regions and multiple algorithms, then scored for confidence instead of presented as fact. Nothing leaves your device. HeartBit runs entirely in the browser, tracking trends over days instead of single sessions, the same way hospitals catch early warning signs. That's the same discipline your compliance team will expect to see.

On-device only Multi-algorithm cross-check Confidence-scored Trend-aware
Side-by-side comparison of a 468-point facial landmark mesh: a wireframe 3D face model labeled '468 points' next to a real face with the same dense triangulated mesh tracked across it.

HeartBit is a research and wellness-tracking tool, not a diagnostic device — an early, low-friction signal to help notice trends worth a closer look, not a replacement for a doctor or a medical-grade monitor.

Clinical & enterprise ready

Rigorous where it counts. Flexible everywhere else.

Privacy-first by design

On-device or in your private cloud, always. Raw video and audio are never kept, only the numbers they produce.

Validated with clinical methodology

Every model is developed using reference-standard devices and reviewed by clinical advisors, to support judgment, never replace it.

Deploys into your existing workflow

Ship as an SDK, a widget, or an API, in the formats your team already uses.

Heart Rate
72bpm
SpO₂
98%
Resp. Rate
16/min
Rhythm
Normal Sinus

Deployment

Deploys wherever your care happens.

Telehealth visits, remote monitoring, clinical research, or point-of-care triage. Our solutions fit into the workflow you already have.

Mobile SDK

Add facial-video and voice capture to your iOS or Android app in a few lines of code.

Web Capture Widget

Embed a browser-based capture flow into your telehealth or patient portal. No install required.

Cloud Inference API

Send signal, get structured results back through a straightforward REST API built for server-side integration.

On-Device / Edge Inference

Run models locally for latency-sensitive or offline-first deployments. Nothing leaves the device.

HL7 FHIR Export

Structured, standards-based output that maps cleanly onto the data model your EHR expects.

EHR-Ready Data Model

Results are structured for clinical review the moment they're generated.

Get in touch

Bring signal-based vitals to your platform.

Partner with our team to explore where contactless signal AI fits your product, your patients, and your regulatory path.

Talk to an expert

Discuss where corbit fits your product and get a tailored technical overview.

Pilot & integration support

Hands-on guidance getting HeartBit or VoiceBit capture into your environment.

Ongoing clinical partnership

Continued validation, model updates, and support as your deployment scales.

FAQ

Frequently asked questions

corbit's signal processing can run on-device or in a private cloud environment you control. Raw video and audio are not retained by default — only the derived measurements are, unless you configure otherwise.
Accuracy varies by signal type and use case, and is validated using reference-standard clinical instruments during model development. Detailed validation methodology and performance data are available to partners under NDA.
A standard smartphone or webcam for HeartBit, and a standard microphone for VoiceBit. No proprietary sensors are required.
Most partners move from pilot to production within a few weeks, depending on the deployment model — SDK, API, or on-device — and your existing infrastructure.
We build and test corbit as rigorously as any clinical product, with a regulatory strategy tailored to each market and use case — from telehealth and remote patient monitoring to hospital-at-home programs. Contact our team for current regulatory status in your region.

corbit

The vitals were always there. Now you can read them.