What is a vocal biomarker here?
Voice and speech features treated as candidate physiological or behavioral signals — structured when present and qualified, not sold as a standalone diagnosis.
Solutions · Endpoints
Physiological endpoints & vocal biomarkers
A physiological data platform for teams that need session-linked voice, autonomic, and wearable signals structured for care, digital therapeutics, and research — not another note generator.
Empty modalities stay empty. Capability follows device coverage and qualification.
Physiological data platform
Structure what the session actually produced
SeroState ingests consented session audio and optional wearable streams, then emits FHIR R4 / QOAX-oriented artifacts when each signal is present and qualified. That is the product stance for endpoints you can defend in a protocol review.
Looking for documentation relief instead? See AI scribe vs session evidence and the therapy documentation burnout guide — same stack, different buyer question.
Modalities
Voice, heart rate, and what wearables allow
- Vocal biomarkers / voice featuresstructured when the recording and processing path support them — not a standalone diagnosis● First-class
- Heart rateprimary autonomic modality when a wearable supplies HR for the session window● Primary
- HRVopportunistic where sampling permits● When present
- EDAonly when the session wearable provides it● When present
- Invented valuesmissing streams do not become placeholders or accuracy claims✕ Not this
What you get
Endpoints you can attach to a record
- Structured exportsession physiology for EHR / research paths (FHIR R4 / QOAX-oriented)
- Qualification honestymissing streams stay missing — no fabricated modalities
- Clinician judgmentSeroState does not replace the decision layer
We do not publish accuracy percentages, guarantee every modality every visit, or name consumer wearables in public copy. Capability follows device coverage and consent.
FAQ
Frequently asked questions
Is this for DTx and trials?
Yes in scope: teams that need physiological endpoints alongside (or instead of) narrative notes. Fit depends on protocol, wearable coverage, and qualification rules.
Do you claim FDA clearance or accuracy %?
No. Public pages do not claim diagnostic accuracy percentages or regulatory clearance.
Is this an AI scribe?
No. See AI scribe vs session evidence for the category split. SeroState does not author the clinical note.