Health and clinical research
Use voice-derived acoustic measurements within approved studies and validation programmes for defined populations and intended uses.
Voice Biomarker API
Integrate voice biomarkers into your application
The Virtuosis API analyzes speech recordings and returns AI-powered voice biomarkers through four analysis families: Wellbeing, Parkinson’s, Alzheimer’s and Communication Coach.
API capabilities and requirements
| Category | Specification |
|---|---|
| Insights | SpecificationStress, anxiety and depression; Parkinson’s-related insights; cognitive insights covering Alzheimer’s disease and mild cognitive impairment; communication analytics. |
| Audio formats | SpecificationWAV, MP3, MP4 or OGG; Base64 in JSON. |
| Audio requirements | SpecificationAt least 30 seconds of free speech; up to 50 MB. |
Built for different workflows
Screening and monitoring
Integrate supported voice biomarkers into workflows that help professionals identify changes and prioritize follow-up.
Digital health products
Add voice analysis to telehealth, remote monitoring and digital-health services through a documented API.
Wellbeing and communication
Support wellbeing and communication applications with speech-derived insights.
How the integration works
1
Review the documentation
Confirm authentication, supported audio inputs and the response fields required by your workflow.
2
Create the account and connect
Create the user account under the organisation associated with your API key, then connect your application through the documented endpoints.
3
Submit and retrieve results
Submit at least 30 seconds of free speech and poll for results until processing is complete.
4
Test the workflow
Evaluate representative recordings and confirm that the returned fields behave as expected before launch.
Responsible access and deployment
Wellbeing insights, including stress, anxiety and mood, and communication insights are available through self-service access. Access to medical insights requires approval by Virtuosis, typically within 48 hours, in accordance with the intended purpose of its CE-marked software as a medical device. Detailed medical outputs are intended to support qualified healthcare professionals and do not replace their clinical judgement.
Voice recordings and derived outputs may contain sensitive personal data. The applicable legal basis and default allocation of data-protection roles are described in our Privacy Policy. A specific hosting region or a different allocation of controller and processor roles can be agreed for each deployment, together with project-specific access, retention and deletion requirements.
Review the evidence and evaluated performance in our clinical validation results.