SomaLoopVerified on hardware
Research-grade data in a consumer band. Screenless, worn for long periods, and the raw data is yours to take.
Why a model company builds hardware
Not comparable means not trainable
Third-party devices each compute a private score: not comparable, not traceable to the raw waveform, and so not usable for training. Our own band yields a consistent sequence that traces back to raw signal.
Why a band and not a ring
Rings are limited by size: mainstream rings last 4 to 8 days, have no ECG and fit only on a finger. A band has room for a side ECG electrode and a larger battery, and can be worn on the wrist or upper arm.
The raw data is yours
Raw PPG and acceleration stream out through the SDK in real time; heart rate, steps, skin temperature and SpO2 are written to device history. That is what makes multimodal labelling possible.
Specifications
| Raw PPG | 50 Hz continuous raw signal, streamed through the SDK (measured) |
|---|---|
| Motion | 3-axis accelerometer (measured; no gyroscope) |
| Skin temperature | Continuous wrist skin temperature (measured; not core temperature) |
| SpO2 | Automatic measurement, each reading written to device history (measured) |
| HRV | Device HRV values (measured); PPG pulse intervals under validation |
| ECG | Single lead, spot check via the side electrode |
| Wear | Wrist or upper arm |
| Battery | 15+ days (manufacturer rating) |
| Water resistance | IP68 (manufacturer rating) |
Items marked "measured" were verified on our own hardware; the rest are manufacturer ratings. Specifications are subject to mass-production sign-off.
How data is acquired
Four things to know before designing a product or a study.
| Heart rate · steps · skin temp | Recorded continuously to device history | Survives time away from the phone; back-fills on reconnection |
|---|---|---|
| SpO2 | Measured automatically, reading by reading | Individual readings, not one average per night |
| Raw PPG · acceleration | Streamed over Bluetooth during capture; not stored on the device | Overnight capture needs a phone or bridge nearby |
| ECG | Spot check via the electrode | Waveform shown, never interpreted |
Data discipline set alongside the hardware
0 = not measured, never reported
The app will never show a fake reading like "SpO2 0%".
Quality flags travel with samples
Motion-corrupted or loose-fit samples are stored but never enter trends or interpretation.
Wrist temperature ≠ body temperature
The field is literally skin_temperature. Until ECG calibration is confirmed, no accuracy or diagnostic claim is made.
In research
Dual-ankle kit
Two SomaLoop bands on the ankles recover left-right contact, impact and other running metrics a single wrist cannot see, at a fraction of the cost of a running power meter.
SomaLoop for BioHackerIn development
A band that opens its raw data and an AI companion that knows you. Your body, your data, your experiments.
iOS closed beta under way, opening in batches; Android in development.
The beta already runs the full band-to-cloud path: data syncs automatically, daily sleep, recovery and load scores are produced, and health monitoring, alerts and notifications are running in a test environment.
Today's body, against your own baseline.
Design mock-ups. The shipping interface may differ.
Who it is for
People already experimenting on themselves who cannot get their own raw data.
Quantified-selfers
Wear several devices and run regular blood, methylation and CGM tests.
N-of-1 experimenters
Supplements, fasting, cold exposure, sleep interventions, all of which they want to measure.
Developers and hackers
Want raw data and an API to write their own analysis.
Opinion leaders
Amplifiers of content and community, natural seed users.
Data: a faithful mirror.
Today
Today's recovery, sleep and load, plus the latest SpO2 and HRV.
Morning report
Last night's sleep and heart-rate curves, first thing every day.
Trends
Weekly and monthly baseline change. Failed-quality samples are removed automatically, and how many is stated honestly.
Export
Take all raw data in one click. No gate, not buried three menus deep.
An AI companion that grows on your data.
Learns your baseline
What your resting heart rate and sleep architecture look like.
Remembers your experiments
Every intervention you ran, from which day to which.
Learns how you respond
How you, specifically, respond to interventions. It changes as data accumulates and knows you better with use.
Core scenario · N-of-1 experiments
"I took supplement X for two weeks. Did my resting heart rate and HRV change attributably?"
Proposing, tracking and reviewing the experiment, the whole loop.
Bring your own data
You already hold blood, genetic and CGM reports. Importing them is planned; scope to be defined.
It comes as four companions: a pace coach, a recovery partner, an action buddy and a data partner. One AI, different voices and ways of being with you. Pick the one on your frequency.
Become an early user
iOS closed beta under way, opening in batches. Tell us your device list, the experiments you are running and the data you want.