Secure cloud data processing
Signal ingestion, valid-window selection, and phenotype inference run in an isolated, access-controlled cloud environment. Compute is separated from identity, so analysis never requires the raw identity of the subject.
Neurosymbolic biometric intelligence
Your cardiovascular system deserves its own normal range.
SaluTests applies neuro-symbolic AI with predefined physiological rules prompts to pulse volume signals after their pre-processing for advanced hemodynamic analysis — moving the health area beyond outdated population-derived reference intervals toward personalized hemodynamic phenotyping. Every reading is judged against your own physiology, not a statistical average of strangers.
HIPAA & GDPR-aligned processing · encrypted end to end
Valid window · signal 0.98Live morphology
- Concordance domains
- 4
- Destructive filters applied
- 0
- Phenotype per person
- 1
How it works
Our scientific edge, in four steps
Most systems destroy the very information they are meant to measure. SaluTests is built the other way around — to preserve the waveform, then reason over it.
- Step 01
Valid signal selection in the cloud
Our trusted & valid windows algorithms scan the raw recording at different stages (locally and in the cloud) and keep only the intervals that are physiologically trustworthy — clean, artifact-free beats with reproducible wave morphology.
Unlike other platforms, we do NOT use destructive filtration that kills pulse wave morphology. Nothing is smoothed, band-passed, or averaged away: the diagnostic shape of every pulse beat survives intact.
Input · raw pulse volume signals - Step 02
Precise pulse wave analysis & phenotype indication
Each retained beat is decomposed into its contour landmarks — upstroke kinetics, systolic peak, dicrotic notch, decay slope — and mapped to an individual hemodynamic phenotype rather than a population percentile.
Output · personal hemodynamic phenotype - Step 03
Physiological cross-concordance assessment
The neurosymbolic AI engine cross-checks the heart, the vessels, and blood volume against one another. Concordant findings confirm the phenotype; discordance is surfaced as a signal in its own right, not discarded as noise.
Domains: Cardiac Work · Its Cost · Vascular Compensation Status · Volaemic (Un)Stressed ReservesBody Space: Central · PeripheralTime Stages: Intervals · Total sessionInterventions: Orthostatic · Clinostatic - Step 04
Resilience, reserves & risk reporting
Health resilience, functional reserves, and emerging risks are quantified, then compiled into an automated recommendations report — reasoned, traceable, and written in physiological or coaching language depending on the selected mode.
Deliverable · automated recommendations report
Technology & security
Built to be trusted with clinical data
Premium health infrastructure is judged on what it refuses to do carelessly. Our architecture treats confidentiality, integrity, and explainability as design constraints.
Encryption in transit and at rest
Every recording is encrypted the moment it leaves the device and stays encrypted in storage. Keys are managed independently of the data they protect, with strict rotation and least-privilege access.
HIPAA & GDPR alignment
Data minimisation, lawful basis, subject access, and deletion workflows are designed in — not bolted on. Processing agreements and audit trails accompany every deployment.
Traceable, auditable inference
Because the engine is neurosymbolic, each conclusion carries the explicit rules and measurements it rests on. Every report can be re-derived and reviewed by a clinician.
- AES-256 at rest
- TLS 1.3 in transit
- Role-based access control
- Full audit logging
Founder & Chief Scientific Director
The science is not outsourced. It is authored here.
Dmitry M. Davydov
Ph.D., D.Sc. — Founder & Chief Scientific Director
Google Scholar publicationsPeer-reviewed record, citations & co-authors
“A normal range built from a crowd tells you how unusual you are. It never tells you how well you are. Phenotyping does.”
- Present
Founder & Chief Scientific Director, SaluTests
Directs the scientific programme behind neurosymbolic hemodynamic phenotyping — from trusted & valid window signal theory through to automated physiogical and health recommendation logic.
- Research
Physiology & autonomic regulation
Decades of peer-reviewed work on hemodynamic regulation, pain and cardiovascular physiology, individual differences, and the interaction between autonomic control and health resilience.
- Method
Individualised reference frameworks
Long-standing critic of population-based normal ranges; developed methodological approaches that establish each person as their own physiological reference, quantifying body and mind working reserves rather than deviations from a group mean.
- Practice
Translational & interdisciplinary collaboration
Extensive international collaboration across physiology, biomedical engineering, and computational modelling, with a publication record spanning experimental studies, reviews, and methodological contributions.
Request live demo
See your own waveform, phenotyped.
Tell us about your clinic, cohort, or research programme. We will arrange a live session using real signals so you can judge the method on evidence, not on claims.
- A 30-minute walkthrough with our scientific team
- Live pulse wave analysis on a sample recording
- Phenotype report structure and integration options