Why a personalized baseline beats a population average. Why your breath has a fingerprint. And why one wellness app cannot be everything to everyone.
THE SCIENCE
The science of awareness, made personal.
Why baselines beat averages
HRV at 45 milliseconds means different things for different bodies. A 30-year-old endurance athlete sitting at 45 ms is recovering poorly. A 55-year-old desk worker at 45 ms is in unusually good autonomic shape.
Population averages are useful for epidemiology. They are useless for telling YOU how you are doing right now.
BAM computes z-scores against your own baseline for every metric — heart rate, HRV, respiratory rate, skin temperature delta. The patented status indicator (green / yellow / orange / red) reflects how far you have moved from your normal — not from anyone else is.
The spectral fingerprint of breath
Voice has formants — concentrated energy in specific frequency bands between 200 Hz and 4 kHz. Nasal breathing is different: broad low-frequency noise, almost all of it below 500 Hz.
During calibration, BAM captures your specific 4-band breath spectrum across 3 inhales and 3 exhales. The detector then matches live mic input against this fingerprint using cosine similarity.
The result: the lung wireframe pulses to YOUR breath. The AI ignores voice, laughter, coughing, or ambient noise. Your nervous system is the data.
4-BAND BREATH SPECTRUM
0–250 Hz65%
250–500 Hz20%
500–1500 Hz10%
1500–4000 Hz5%
Typical nasal breath fingerprint. Voice would invert this — energy concentrated in 500–4000 Hz formant bands.
The patented progress tool
The color-coded status indicator (the heart of US Patent 12,109,041) is a clinical-style triage system applied to consumer wellness. Green / yellow / orange / red is universal because everyone has used a traffic light.
But the math underneath is specific to you. Heart rate uses z-scored baseline. SpO₂ uses absolute clinical thresholds (under 92% = red, regardless of history). Breath rate combines both. Each metric gets its own evaluator — and the worst color across all of them becomes the overall status that drives the AI.
Why three modes, not one
Stress is not a single experience. The biometric signature of winding down at home differs from regulating during a difficult conversation, which differs from pacing a 5K.
An app that tries to be everything to everyone ends up generic. We took the opposite approach: three intentional modes that bring distinct technique libraries to the same physiological data.
Slow breathing, body scans, parasympathetic activation. For homeostasis.
Somatic inquiry, affect labeling, pendulation. For embodied insight.
HR zones, breath-pace matching, overtraining detection. For training.
Honest limits
BAM is a wellness tool, not a medical device. We do not detect arrhythmias, predict cardiac events, diagnose mental-health conditions, or replace clinical care. The patent claims a method of monitoring and training physiological responses to stimulation — not diagnosis.
If you have a medical concern, see a clinician. If you are in mental-health crisis, contact a crisis line or emergency services. BAM exists alongside qualified care — never instead of it.
See the method in action.
The science is the foundation. The Vision Pro app is where it lives.