Ellipsis Health: Turning Everyday Conversations Into Early Warnings for Mental Health
What is Ellipsis Health? Ellipsis Health combines clinically validated vocal biomarkers (short voice recordings analysed by ML) with an enterprise AI care-manager called Sage to screen, monitor and autonomously engage patients at scale. The platform is designed for population screening, care-management call workflows, telehealth follow-ups and chronic-care outreach — surfacing depression/anxiety signals, prioritising high-risk callers, […]
Feature Categories
What is Ellipsis Health?
Ellipsis Health combines clinically validated vocal biomarkers (short voice recordings analysed by ML) with an enterprise AI care-manager called Sage to screen, monitor and autonomously engage patients at scale. The platform is designed for population screening, care-management call workflows, telehealth follow-ups and chronic-care outreach — surfacing depression/anxiety signals, prioritising high-risk callers, and escalating to human clinicians when needed.
Ellipsis positions its tech as a low-friction, privacy-guarded supplement to PHQ/GAD screening, enabling earlier detection and continuous monitoring while reducing administrative burden for health plans and systems.
Why Leading Healthcare Teams Trust Ellipsis Health
- Trusted by leading healthcare institutions and operates as a cloud-based solution that is highly scalable and platform agnostic
- Secured $45 million Series A funding led by Salesforce, Khosla Ventures, and CVS Health Ventures in 2025, demonstrating strong investor confidence
- Total funding of $75 million from tier-one investors including Vinod Khosla, Marc Benioff, ICONIQ Capital, and other prominent venture firms
- Uses voice as a biomarker to identify, measure, and monitor stress, anxiety, and depression at scale through natural speech analysis
- Provides objective clinical decision support tool through partnerships with providers, payers, and digital health companies
- Founded in 2017 with 39 total investors and operates in the highly regulated mental healthcare space
- Focuses specifically on mental health care utilizing voice biomarkers within the healthcare technology sector
- Recently launched Sage, their emotionally intelligent AI Care Manager product
Top 3 Pain Points Ellipsis Health Fixes in Healthcare
| Problem | How Ellipsis Health Solves It |
|---|---|
| 1. Missed or delayed detection of depression/anxiety | Uses AI voice biomarkers to identify mental-health signals early and continuously. |
| 2. Overburdened clinicians and care managers | Automates routine screening, triage, and follow-up calls, freeing clinicians for higher-value tasks. |
| 3. Gaps in population-level mental health monitoring | Scales across large patient groups, integrating into call centers and care-management workflows. |
Feature Category Summary: Ellipsis Health
| Feature Category | Summary | Association (YES, NO, NA) |
|---|---|---|
| Regulatory-Ready | Ellipsis Health highlights SOC 2 Type II, GDPR, and HIPAA compliance, and notes it is working toward HITRUST certification, but there is no mention of FDA or EMA clearance, 21 CFR Part 11/Annex 11 validation, or GxP audit‑trail/CSV documentation; the tools are positioned as clinical decision support and care‑management aids, not as cleared medical devices. No public documentation found that the platform is validated as a regulatory‑ready GxP/Part‑11 system. | NA |
| Clinical Trial Support | Use cases focus on screening and monitoring mental‑health symptom severity, augmenting PHQ‑8/9 and GAD‑7, and automating outreach through Sage for enrollment, assessments, adherence checks, and post‑discharge follow‑up. There is no indication of functionality for trial protocol design, subject recruitment logistics, trial endpoint monitoring, or submission reporting. | NA |
| Supply Chain & Quality | The platform analyzes speech to assess mental health and drive care‑management workflows; no capabilities are described for pharmaceutical manufacturing integrity, batch QA, serialization, or counterfeit‑medicine detection. No public documentation found for supply‑chain or manufacturing‑quality support. | NA |
| Efficiency & Cost-Saving | Ellipsis positions Sage as an “emotionally intelligent AI care manager” that performs fully autonomous multilingual phone outreach for enrollment, assessments, coordination, and follow‑up, explicitly aiming to “immediately expand capacity,” “reduce costs,” and “supercharge existing workflows” for providers, payers, and care‑management organizations. Research posts also emphasize that automated pre‑visit screening and continuous voice‑based monitoring free clinicians’ time and make patient encounters more efficient. | YES |
| Scalable / Enterprise-Grade | Ellipsis describes its SaaS platform as HIPAA and SOC 2 Type II compliant with FHIR‑based data standards and API availability, and customer lists include large organizations such as Highmark Health. Series A funding of $45M led by Salesforce, Khosla Ventures, and CVS Ventures is explicitly intended to scale Sage adoption across providers, payers, and care‑management organizations, indicating an enterprise‑grade, healthcare‑focused deployment model (though not in pharma/biotech). | YES |
| HIPAA Compliant | Security materials state that Ellipsis Health adheres to HIPAA and SOC 2 Type II, with end‑to‑end encryption, regular third‑party security audits, and compliant handling of sensitive mental‑health data. Company descriptions reiterate “HIPAA and SOC2 Type 2 compliant infrastructure” as a core attribute of Sage, which is explicit evidence of HIPAA‑aligned operation. | YES |
| Clinically Validated | Ellipsis repeatedly describes its voice biomarker as “clinically validated,” citing peer‑reviewed work, including an IEEE study showing the algorithm is a “viable technique” accepted by patients of different ages and races, and a feasibility study where the Ellipsis app successfully screened for depression and anxiety across age groups by comparing its scores with PHQ‑9 and GAD‑7. University of Denver and University of Michigan research partnerships are actively validating the tool in adolescents, directly comparing Ellipsis scores with standard questionnaires, reinforcing clinical validation for symptom‑severity assessment use. | YES |
| EHR Integration | Ellipsis states that its voice‑biomarker assessments can be embedded in telemedicine calls, digital front doors, and apps, and HealthAidb notes use of FHIR and API‑based integration into care workflows, but specific named EHR integrations (Epic, Cerner, etc.) are not detailed. No public documentation found that explicitly confirms direct integration with particular EHR vendors or in‑EHR launch workflows. | NA |
| Explainable AI | Company materials mention “transparent AI decision‑making processes,” and research posts describe that the system analyzes both semantic content and acoustic features and compares voice‑derived scores with PHQ/GAD measures, but there is no description of end‑user features that expose model rationales, feature‑importance, or per‑prediction explanations. Transparency is framed in general terms, not as concrete explainable‑AI tooling accessible to clinicians. | NA |
| Real-Time Analytics | Ellipsis technology analyzes ≈60‑second voice recordings and can be “layered seamlessly onto calls between a provider and a patient,” enabling near real‑time assessment of anxiety/depression severity during live interactions or routine check‑ups. While processing is encounter‑based rather than streaming physiologic data, it does provide on‑the‑fly symptom‑severity analytics within clinical workflows. | YES |
| Bias Detection | Ellipsis explicitly addresses bias: its safety and data‑protection FAQ describes deliberate collection of hundreds of thousands of recordings from diverse demographics and asks, “How does Ellipsis Health reduce biases in its AI models?”, outlining steps to build representative datasets to avoid bias. Research notes that its large, diverse voice databases and peer‑reviewed IEEE study show acceptance and performance across ages and races, which constitutes explicit bias‑mitigation and monitoring at the data and model‑evaluation level, even if not exposed as a separate product module. | YES |
| Ethical Safeguards | Ellipsis repeatedly emphasizes privacy, consent, and data security “at the forefront”, noting adherence to HIPAA, SOC 2, GDPR, and work toward HITRUST, alongside clinical oversight of all AI operations, continuous validation, and monitoring. Company statements and independent coverage stress that the tool does not provide a diagnosis, but supports clinicians by assessing symptom severity and augmenting PHQ/GAD surveys, keeping human clinicians in the loop and positioning Sage as assistive, not autonomous clinical decision‑maker. | YES |
Risks & Limitations: Ellipsis Health
-
Audio quality & environment sensitivity: background noise, low-fidelity calls, and telephony compression reduce signal quality and diagnostic accuracy.
-
Population & language bias: acoustic and linguistic models may underperform on accents, dialects and languages not represented in training data — local validation and re-calibration are essential.
-
Regulatory / clinical acceptance: voice-biomarker outputs are advisory; clinical adoption requires documented pathways, validation against local cohorts, and medico-legal clarity.
-
False positives & workflow burden: without tuned PPV thresholds, triage flags can overwhelm behavioural-health teams; set operational thresholds and volume caps in pilots.
Tags
Share This AI Tool
Get a neutral, no obligation view of whether this AI tool fits your portfolio
Stephen
Founder of HealthyData.Science · 20+ years in life sciences compliance & software validation · MSc in Data Science & Artificial Intelligence.