The Bio-intelligence layer for wearable health.

Rhythmic AI for engineering and health. Real-time biosignal processing across PPG, ECG, EEG, and IMU, delivered as a simple API.

Contact usRead the docs
POST /v1/ppg/afibLive · 128 Hz
Heart rate
72bpm
AFib
sinus
Confidence
0.97
{
  "afib": false,
  "hr_bpm": 72,
  "confidence": 0.97,
  "latency_ms": 184
}
The problem

The bottleneck in wearable health isn't hardware. It's the algorithm layer.

01
$350B
Digital health market
The bottleneck isn't hardware. PPG sensors cost <$1. Chips are commodity. The bottleneck is algorithm validation.
02
<5K
Biosignal + AI PhDs
Fewer than five thousand in the world. Companies wait 12–18 months to hire one, if they can find one.
03
$1M+
Reinvented per company
Every wearable startup rebuilds the same AFib, HRV, SpO₂ stack from scratch. Most of it is already solved.
04
58%
Withings AFib sensitivity
And most still fail. Devices ship with algorithms that miss nearly half the events they claim to detect.
Solutions

Four kits. One biosignal stack.

Each kit is a bundle of production-ready algorithms you drop into your wearable. Validated, confidence-scored, and production-ready. No PhDs required.

Kit 01

Heart Kit

The cardiac layer, built on a single PPG stream. Rhythm, zones, recovery, and events in one bundle.

Primary signalPPG
Modules5
AFib sens.>95%
  • 01
    Rhythm intelligence
    Continuous beat-to-beat readings across rest, activity, and recovery. Not just headline averages, but a live picture of how the heart is actually behaving.
  • 02
    Performance windows
    Effort and recovery zones framed against the wearer's own baseline. Meaningful context for athletes, and a useful anchor for clinical pathways.
  • 03
    Event detection
    Surfaces cardiac irregularities worth a second look, each with a confidence score. Your app decides how to present them. We keep the flags honest.
Talk to us about the Heart Kit→
01 / 04Heart Kit·auto-playing
How it works

Two WebSockets. Real-time results.

Your device streams raw samples in. Your app subscribes to structured insights out. Confidence scores on every metric, typically under a second end-to-end.

  1. 01
    Register your device
    In the dashboard, define your hardware profile (sensors and sample rates). You get a device_model_id to bake into firmware.
  2. 02
    Mint an API key
    Scoped to your account, prefixed raeh_. Ship one with your firmware, or mint per-user keys from your backend.
  3. 03
    Stream raw samples
    Your firmware opens /stream/ingest over WebSocket and pushes binary PPG frames at the sensor's native rate.
  4. 04
    Receive structured JSON
    Your app opens /stream/subscribe and receives real-time insights (HR, SpO₂, RR) with a confidence score on every value.
subscribe.py
import asyncio, json, websockets

async def main():
    url = "wss://api.raeh.io/stream/subscribe?api_key=raeh_..."
    async with websockets.connect(url) as ws:
        await ws.recv() # subscribe ack
        async for msg in ws:
            m = json.loads(msg)
            print(m["type"], m["value"], m["unit"])

# hr 72 bpm
# spo2 98 %
# rr 15 brpm
Get started

Build on truth.
Ship with confidence.

Whether you're building a smartwatch, a clinical monitor, or the next generation of health platforms, RAEH provides the algorithm foundation you can trust.

Build with us