Three steps: scan a real biological signal, understand it as one honest score, act on guidance from Gabriel and the specialists.
A 20–25 second camera scan measures pulse-derived signals (heart rate, heart-rate variability, breathing rate, pulse-wave shape). You can sync a wearable instead or in addition, and upload lab results manually. Signal quality is reported with every scan; low-quality scans are labelled, not silently used.
Readings are combined into the Human Efficiency Score (HES), a 0–100 composite across five pillars — Nutrition & Metabolism, Recovery & Rhythm, Sleep, Training & Body and Mental & Resilience — with Nutrition & Metabolism weighted at 25% and Recovery & Rhythm at 22%. The score is shown as a range, not a single verdict, when too few pillars have real data.
Gabriel and the specialists translate the score into concrete, non-medical suggestions: training load, sleep window, meals, focus blocks. Once a week, a single synthesis meeting merges every specialist's view into one briefing instead of seven separate messages.
You can export or delete your data at any time. Biomarker values are converted into qualitative phrasing before any AI companion sees them, so raw clinical numbers are never passed into an AI prompt.
The scan is remote photoplethysmography (rPPG): the camera detects the minute colour changes in your skin caused by each heartbeat. From that pulse waveform the engine derives heart rate, heart-rate variability (SDNN and RMSSD), breathing rate, and shape features of the pulse wave itself. Exposure and ISO are locked during the scan so the signal is not corrupted by the phone's own auto-adjustments, and both lenses can be used in sequence for a more robust reading.
Every scan reports its own quality. The engine uses a triple consensus — frequency analysis, autocorrelation and inter-beat-interval median — and a scan that fails to agree with itself is labelled low quality rather than quietly averaged into your history. This matters more than headline accuracy: a wellness score built on unmarked bad data is worse than no score at all.
You are not expected to scan constantly. Optional wearable sync (via your phone's health platform, plus connected rings and watches) fills the gaps with continuous data. Sleep is tracked as its own cycle, with a protocol that adapts to your work schedule rather than assuming a fixed bedtime. Nutrition can be logged by barcode. None of it is mandatory — the score degrades gracefully when a source is missing instead of inventing a value.
Gabriel and the six specialists read one shared wellness state. Each specialist sees only the signals its own domain needs, so the nutrition companion never receives your training history and the social companion never receives lab values. Biomarkers are converted into qualitative phrasing before any AI companion sees them — 'toward the higher end of your own recent range', never a raw clinical number. Every included and excluded signal is logged, per request.
Once a week the specialists' views are merged into a single briefing rather than seven separate notifications. It leads with what actually changed, not with a recap of what you already know, and it is explicit about what it does not know. This is the moment the product is designed around.
Quantum Khor is a wellness companion, not a medical device. It does not diagnose, treat or prevent any condition.