Voxwell.ai
Voice AI for mental health 60-second screening

Detect risk sooner. Using just a voice sample.

Voxwell.ai leverages advanced AI and vocal biomarkers to transform a short, natural speech sample into an objective, privacy-preserving signal for depression and anxiety risk — enabling earlier screening at scale across care pathways and populations.

Get in touch Investor brief
60s
Voice sample
API
Drop-in integration
Edge
Privacy by design
Scale
Population ready

Speak naturally for 60 seconds. Voxwell’s AI converts micro-acoustic patterns into risk signals and a clear next-step recommendation.

Input
Free speech
Signal
Objective score
Action
Triage & refer
Works in any language & accent*

*Model performance varies by cohort; clinical validation in progress.

About Voxwell

Voxwell is building the missing vital sign for mental health. Our platform analyzes subtle, non-linguistic patterns in natural speech to flag elevated risk for depression and anxiety in under a minute. Designed for health systems, digital clinics, and employers, Voxwell enables proactive outreach, smarter triage, and measurable impact on outcomes and cost.

  • Objective, biomarker signal that complements clinical judgment.
  • Privacy-preserving features engineered to minimize PII exposure.
  • API-first: embed in telehealth, care management, or employer apps.

At a glance

60s
screening time
SDK / API
easy integration
Edge-ready
on-device option
HIPAA
alignment

Problem

Millions suffer in silence due to the stigma, cost, and inaccessibility of traditional mental health screening. The current system is reactive, subjective, and fails to reach those in need before a crisis occurs.

1 in 5

adults experience mental illness each year, yet less than half receive treatment.

7-11 Years

is the average delay between symptom onset and intervention.

$1 Trillion

is the estimated annual global economic loss due to depression and anxiety.

Vision

Make early mental health screening as simple and universal as checking a pulse. By unlocking voice as a vital sign, Voxwell helps organizations identify risk earlier, close care gaps, and direct people to the right level of support — at population scale.

Access

Frictionless screening within existing touchpoints — calls, apps, or kiosks.

Equity

Language-agnostic signal reduces bias from literacy and self-reporting.

Impact

Earlier identification enables timely intervention and cost avoidance.

Technology

Built on years of research in speech signal processing and machine learning, Voxwell's technology converts micro-acoustic dynamics (e.g., prosody, phonation and articulatory features) into robust embeddings and risk scores using proprietary models. Our approach emphasizes acoustic features to protect privacy while maintaining signal quality across languages and accents.

  • Objective & Data-Driven: Replaces subjective questionnaires with quantifiable metrics. Model ensemble: classical DSP + deep representation learning. Calibration toolkit for specific populations and microphones.
  • Scalable & Accessible: Deployable on any device with a microphone, enabling population-level screening. Cloud, on-prem, or on-device inference paths.
  • Secure & Compliant: End-to-end encryption and HIPAA-compliant architecture ensure data privacy. Non-linguistic feature focus to minimize PHI exposure.

How it works

  1. 1 User speaks naturally for ~60 seconds into a recording device.
  2. 2 Audio is converted into secure spectral features on-device or server-side.
  3. 3 Voxwell models output a risk score and recommended next step.
  4. 4 Integrations trigger triage, referrals, or supportive content.
Clinical roadmap. IRB-guided studies underway to validate sensitivity/specificity across diverse cohorts.

Solutions

Whether you run a hospital call center, a digital behavioral health program, or an employer benefits platform, Voxwell fits in fast.

Health Systems

Embed screening in the triage and intake processes to provide an objective support, replacing the current subjective questionnaires.

Clinician dashboard
Digital Health

Integrate into virtual visits to improve time-to-care and enhance measurement-based care.

Mobile SDK
Employers & Payers

Identify rising-risk members and direct them to EAP or covered services sooner.

Population analytics
Delivery options
  • Web & mobile SDKs
  • Telephony (IVR/call-center)
  • Batch analytics
Integration
  • FHIR / HL7 hooks
  • EMR & CRM connectors
  • SFTP / REST APIs
Outputs
  • Risk score (0–100) with confidence band
  • Trend over time
  • Next-step recommendation
Talk to sales See FAQs

Evidence & Traction

  • Backed by peer-reviewed groundwork in vocal biomarkers for mental health.
  • Awarded one of Top 3 of MerageNEXT 2025.
  • A pilot with a gov. regulatory body.
  • Prospective studies launching with healthcare partners (IRB).
  • Pilot programs available for clinics, hospitals, telehealth, and employer benefits.

Example success metrics

↓ TtC
Faster time-to-care
+ID
More at-risk identified
↓ Util.
Avoidable ER visits
ROI
Demonstrable savings

*Actual metrics reported per-deployment; claim-safe placeholders here.

Exploring Collaboration

Voxwell is currently in discussion with leading healthcare and research institutions to advance clinical validation and integration.

Teva

Teva Pharmaceuticals

Sheba Medical Center

Sheba Medical Center

Resilience Accelerator

Resilience Accelerator

Merage Foundation

Merage Foundation

Wolfson medical center

Wolfson Medical Center

Trust, Security & Ethics

Privacy by design

Feature extraction minimizes exposure to raw content; optional on-device processing.

Security

Encryption in transit/at rest, scoped retention, role-based access, audit logs.

Responsible AI

Bias testing, cohort calibration, human-in-the-loop triage, and clear model labeling.

Team

Multidisciplinary founders across AI, speech science, and healthcare operations.

Prof. Julie Cwikel
Co-founder & Social epidemiologist

Director, Center for Women's Health Studies and Promotion at Ben-Gurion University.

Dr. Dan Vilenchik
Co-founder & CTO

Expert in ML, School of Electrical & Computer Engineering, Ben-Gurion University.

Efrat Kaul Granot
Co-founder & CEO

Attorney & entrepreneur with 14+ years in high-tech across BD, marketing and ops.

Ruslan Sergienko
Co-founder & CIO

Data scientist, software development & IT expert with extensive clinical research experience.

Why invest in Voxwell

  • Category momentum: Voice biomarkers are emerging as a new vital sign, with growing clinical validation and adoption by health systems.
  • Defensible tech: Non-linguistic feature focus and calibration toolkit create privacy and performance advantages.
  • Go-to-market fit: API-first product plugs into high-volume touchpoints (call centers, virtual care, employers).
  • Impact thesis: Earlier identification drives measurable outcomes, adherence, and cost savings.

Round & use of funds

  • • Clinical validation across diverse cohorts
  • • On-device inference & SDKs
  • • Health-system integrations & payer pilots
  • • Regulatory & quality (ISO 13485 roadmap)
Request investor deck

FAQs

Is Voxwell a diagnostic tool?

No. Voxwell provides an objective screening signal that complements existing tools and clinical judgment. It helps prioritize outreach and route people to appropriate care.

What languages are supported?

Our models focus on non-linguistic features and generalize across languages and accents. We calibrate per population/microphone for best performance.

How is privacy protected?

We minimize exposure to raw content via feature extraction, support on-device inference, and apply encryption, scoped retention, and access controls.

How fast can we go live?

Most pilots start in 4–6 weeks depending on integration depth, data protection agreements, and cohort calibration.

Get in touch

Whether you're a healthcare provider, a potential partner, or an investor, we'd love to hear from you. Reach out to learn more about our pilot programs and partnership opportunities.