Videra Health

The models are the product rather than a layer on top of it. Every product in the line runs model inference as its core operation: assessment scoring from video and voice, ambient session documentation, conversational intake, speaker separation in group therapy, and clinician coaching derived from communication patterns. Strip the models out and what remains is a video capture form.

What lifts this above the platform vendors graded lower elsewhere in this index is that the intelligence is proprietary and purpose built rather than assembled. The company trained its own instruments, PHQ-V, GAD-V, PCL-V and EPDS-V, as video native versions of the standard written scales, on a stated corpus of roughly 400,000 training samples and 50,000 hours of clinical data. Those are not general models pointed at a health problem; they are instruments built for one clinical domain and published as such.

A workflow layer exists and is substantial. Intake paperwork, scheduling capture and inbound inquiry routing would retain some value without the models. Graded A rather than qualified because that workflow exists to deliver model inference to a patient and return a result to a clinician, which is a different relationship from the coordination and services platforms where intelligence is one component among several.

The boundary is stated clearly and repeatedly, and the escalation behind it is undefined.

What is stated is unambiguous and consistent across pages. Models provide decision support and never replace clinical judgment, critical decisions remain with qualified clinicians, and results and alerts surface in a provider dashboard and the record system rather than triggering autonomous action. Patient facing material draws the scope boundary honestly, telling patients not to use the platform during a medical or mental health emergency and to call emergency services instead. Stating what a product must not be used for is a discipline this index rewards, and several behavioral health vendors examined recently omit it.

The gap is what happens between the model firing and a human seeing it. Assessments are asynchronous by design, completed by patients on their own devices at whatever hour they choose, and the instrument set includes the Columbia Suicide Severity Rating Scale. That combination produces a foreseeable scenario: a patient completes a check in overnight, the model registers severe risk, and nothing published states who is notified, through what channel, within what time, or whether any coverage exists outside business hours. There is no staffed crisis capability of the kind one other vendor in this lane operates, so escalation rests entirely on the customer's own staffing, and nothing describes what the customer is advised to have in place.

No alerting threshold is published either, so the operating point deciding which patients reach a human is unknown.

Ask for the escalation protocol on a severe risk result, the notification time, the after hours arrangement, and the threshold that triggers it.

The most specific model disclosure in the behavioral health lane, with one gap that matters clinically far more than its size suggests.

Disclosure runs well past the usual. Three modalities are described at feature level rather than by label: linguistic analysis naming absolutist language patterns, word complexity and sentiment; acoustic analysis naming speech rate, pause duration, pitch variation and articulation; and computer vision naming facial movement, microexpressions and affective state. Five models are set out in a table with clinical application, discrimination and modality against each. Training corpus size is quantified at roughly 400,000 samples, 50,000 hours of clinical data and 150,000 research participants. Five peer reviewed papers document the methods, so a technical reader can inspect how the instruments were built rather than accepting a summary on trust.

The gap is the operating point. Area under the curve is threshold independent and describes ranking, not decisions. What a deployed screening tool actually does is apply a cut off, and the clinically relevant numbers are sensitivity and specificity at that cut off plus positive predictive value at the prevalence of the screened population. None of those is published for any model. A tool reporting 0.90 can be tuned to miss most true positives or to flood a clinic with false ones, and the published figure does not distinguish those two deployments.

Model cards are stated to exist and are gated behind a request, which is where an operating point would normally live. Architecture is not described, and any dependence on third party base models is left unaddressed.

Graded A because what is published is quantified, checkable and externally reviewed, which is the bar this axis sets. Ask for sensitivity, specificity and positive predictive value at the deployed threshold for each instrument.

The software supply chain is documented well and the model supply chain is not documented at all.

On the software side there is real substance, which is why this sits above the vendors graded lower on this axis. Amazon Web Services is named as the infrastructure provider. Dependency and third party library analysis is described with scoring criteria that block vulnerable components from release. Container images are assessed with a passing score required before deployment. Those are genuine supply chain controls, described specifically enough that a buyer can ask about them.

On the model side nothing is disclosed. No base or foundation model is named, and no statement addresses whether any third party model is used at all. That absence is conspicuous rather than neutral, because two products in the line are the kind rarely built from scratch: ambient session documentation and a conversational agent handling inbound patient inquiries around the clock. If a general purpose language model sits behind either, patient session audio or inquiry content passes to a party the customer has not been told about; if none does, saying so plainly would be a straightforward differentiator.

No sub processor register was located. No transcription provider is named, despite transcribed speech being an explicit input to the language models. No position is published on whether customer data contributes to model development.

The proprietary instruments are a separate matter and are documented well; this axis concerns what sits underneath and around them.

Ask for the sub processor register, whether any third party model or transcription service processes patient audio, and what the data processing terms with each provider permit.

The strongest published evidence base in the behavioral health lane of this index, with one structural qualification that belongs beside it.

Five peer reviewed papers are listed with full citations and working links, across journals that carry real review: npj Digital Medicine in the Nature portfolio in 2025, JMIR Mental Health in 2026, the Journal of Clinical Psychiatry in 2025, Women's Health Reports in 2026, and Applied Sciences in 2022. The JMIR paper is a cross dataset evaluation of the tardive dyskinesia model, testing performance against data the model was not developed on. External validation of that kind is the strongest form available short of a trial, and no other behavioral health record here has one. Discrimination is published per model rather than described, ranging from 0.85 to 0.92 area under the curve across five instruments on stated independent test sets, with research scale given as roughly 150,000 participants.

Commercial evidence is separate and also real. More than 300 facilities, over one million patient interactions, a named deployment built with Discovery Behavioral Health and Brigham and Women's Hospital, and federal consortium work on service member suicide prevention. Federal procurement diligence is scrutiny of the approach rather than of the marketing.

The qualification is authorship. All five papers carry company staff as authors, and the scientific advisor named on the company site is a co author on both tardive dyskinesia papers. Peer review is genuine and vendor authored research is normal in this field, but there is no independent replication, and the flagship movement disorder validation carries a related party.

The second gap is what gets measured. Every paper validates model accuracy against a clinical reference. None measures whether using the platform improves patient outcomes. Marketing figures of 64 percent fewer crisis driven interventions and 85 percent faster engagement carry no citation, site or method. Ask for the outcome study, not another accuracy paper.

Handling controls are described in more operational detail than almost any record in this index, and the single question that matters most for this vendor goes unanswered.

The detail is specific rather than gestural. AES-256 and TLS encryption in transit and at rest, applied across disk, application and database layers, with key access restricted, key sharing prohibited and key management reviewed annually. Access runs on least privilege with SAML 2.0 single sign on and customer configurable two factor authentication, while multi factor authentication is mandatory for employees through a centralised directory. Networks sit in isolated virtual private clouds behind security groups and firewalls. Customer data is segregated by unique identifier. Retention follows a data classification, secure disposal is described for physical media, and deletion is available on written request. Security event and audit logs are collected and continuously monitored with integrity protection. Patients are told when artificial intelligence is used, stated as an explicit principle rather than buried in terms.

The unanswered question is provenance. The company states a training corpus of roughly 400,000 samples, 50,000 hours of clinical data and 150,000 research participants, and nowhere states whether customer patient recordings contribute to model development, under what consent, or whether a customer can decline. The material at issue is video and audio of a patient's face and voice discussing their mental health, which is among the most identifiable and most sensitive data any vendor in this index holds. A recording cannot be de identified the way a claim line can.

One pre emptive note: further security controls cannot move this grade. Only a published position on whether customer recordings train models, and how consent is obtained, will.

A better than typical posture carrying one notable strength and one wording error.

The strength is specific and rare in this lane. The company publishes a redisclosure consent notice as a standing page alongside its privacy policy. Redisclosure consent is the mechanism governing substance use disorder treatment records under the separate federal confidentiality regime that applies to them, and a behavioral health platform reaching into addiction treatment settings sits squarely inside that regime. Publishing the notice indicates it has been engaged rather than overlooked, which is more than the other behavioral health records here demonstrate. Patient facing terms also state that health history is shared only with the participating care team and not with an employer, and that removal can be requested with a stated processing window of five to seven business days.

The wording error is the one this index records repeatedly. Security material states that the company holds industry standard certifications and then lists SOC 2 Type II and HIPAA compliance together. There is no HIPAA certification. It is a regulatory obligation with no certifying body, and placing it beside an audited attestation invites a buyer to credit two externally verified things where only one exists.

Business associate agreement handling is not addressed anywhere. No template, no execution requirement, no statement of which entity contracts, and no subcontractor flow down position was located.

Ask for the business associate agreement template and how substance use records are segregated in practice.

A public trust center carrying the actual report, which is the top of this axis and rare across this index.

The trust center is live at a public subdomain and is stated to hold the SOC 2 report, HIPAA documentation and live security monitoring. That is a materially different posture from a badge on a marketing page, because a buyer can begin diligence before speaking to sales, which is precisely the openness this index exists to reward. An executive summary with sensitive material redacted is offered under agreement, and a security contact address is published.

The underlying programme is described rather than asserted. SOC 2 Type II is held, annual penetration testing is performed by independent third party agencies, and the secure development lifecycle is set out in unusual detail: automated security testing before release, dependency and third party library scanning with scoring criteria that block vulnerable components, dynamic application security testing, container image assessment requiring a passing score before deployment, and enforced source control standards covering role based access, segregation of duties, branch protection and secrets management. Incident response procedures, personnel background checks and continuing security training are all described.

Three things remain worth asking despite the A. The audit period and the assessor are not named publicly. No health specific certification is held, which one other vendor in this lane does hold and which maps a control set onto health privacy requirements directly. And no vulnerability disclosure policy or coordinated reporting channel was located, the one common component absent from an otherwise complete programme.

Graded A on the strength of a publicly reachable report and a documented programme. Ask for the audit period, the assessor and the report scope.

A credential is being presented as something it is not, while the underlying position is probably defensible.

The company describes itself as an FDA registered digital health platform, repeats the phrase in every press release, and displays an FDA Registered badge in the site footer beside its HIPAA and SOC 2 marks. Establishment registration and device listing is an annual filing obligation. It records that a firm exists and what it lists, involves no review of safety or effectiveness, and confers neither clearance nor approval. The agency itself states that registration does not denote approval and that using it to suggest endorsement is improper. Set in a row of audited trust marks, a buyer will read it as a third party quality signal, which is the one thing it is not. No clearance, De Novo grant or approval was located for any product.

The substantive position is stronger than the presentation. Screening tools returning a score to a clinician who independently decides can fall within clinical decision support treatment, and the company's framing of decision support that never replaces clinical judgment is the correct framing for that path.

Two things keep it uncertain. The decision support exemption generally turns on whether a clinician can independently review the basis of a recommendation, and a score derived from facial microexpression and vocal acoustic analysis is not reviewable the way a rules based flag is; a clinician cannot inspect why a pitch variation pattern produced a depression score. And detecting tardive dyskinesia from video is closer to identifying a physical sign than to summarising patient reported answers.

This is the third regulatory misdescription recorded in this lane, after a designation reported as a clearance and a funding failure reported as an agency rejection. Ask for the written device determination and the regulatory rationale covering each model.

A published ethical framework with no measurement behind it, set against a bias risk profile close to the worst case for this modality.

The framework exists and is more than boilerplate in places. Seven principles cover patient centred care, transparency, equity and accessibility, accuracy, ethical data use and human oversight, and the equity principle explicitly claims diverse datasets used to minimise bias and ensure equitable treatment across patient demographics. Telling patients when artificial intelligence is used is a real commitment.

Nothing measures any of it. No fairness or bias testing, no performance broken down by any subgroup, no calibration data, no drift monitoring and no external audit was located. Model cards are stated to exist but only on request, so the one artefact that would normally carry subgroup performance is gated.

The risk here is not generic. Facial analysis carries well documented differential error rates by skin tone, and acoustic analysis by accent, dialect, sex and age. Beyond demographics the signal itself is confounded in this population: affective expression norms vary by culture, and several conditions the platform is used around alter the exact signals being measured. Antipsychotic medication and tardive dyskinesia itself change facial movement, which is simultaneously the input to the movement disorder model and a confound for the affect models running on the same video. Flat affect is both a clinical feature and a measurement artefact.

An aggregate area under the curve of 0.89 is entirely compatible with materially worse performance inside a subgroup. The published figures do not answer that question and cannot.

Ask for discrimination broken down by race, skin tone, sex, age and primary language, and for calibration within each group.

Nothing allocates responsibility, in the clinical domain where the consequence of a miss is gravest.

A dedicated pass located no service level agreement, no accuracy warranty, no performance guarantee, no indemnity, no remediation commitment and no stated recourse if a model output contributes to harm. Published terms of use are patient and employer facing rather than a customer contract, and nothing describes what a health system's agreement contains.

The published accuracy figures do not substitute for this and should not be read as though they do. An area under the curve is a research statistic, not a commitment, and because no operating threshold is published there is no stated false negative rate for any instrument. For a platform screening with the Columbia Suicide Severity Rating Scale, the false negative rate is the number that matters, and it is unavailable in both senses: neither disclosed nor warranted.

The allocation question is concrete rather than theoretical. A patient completes an overnight check in, the model does not flag deteriorating risk, no clinician is alerted, and harm follows. Responsibility then sits somewhere between an undisclosed model threshold, the customer's own escalation staffing and a vendor carrying no published commitment, and nothing in public material indicates where. This is the first question a health system's counsel should ask, and there is currently no published answer.

One pre emptive note: further publications, higher reported discrimination or additional customer counts cannot move this grade. Only a contractual term, or a published operating point carrying a stated false negative rate, will.

Seven systems named, with no standard and no direction of flow.

The naming is real capability evidence rather than a general claim of openness. Kipu, Alleva, BestNotes and BlueStep are behavioral health, addiction treatment and senior living record systems, which is exactly where this vendor's customers sit, and Epic and Cerner cover the acute market. Zoom appears alongside them as a session capture path. A vendor naming the niche systems as well as the two large ones has usually built to them, and the company states that connecting once to the record system unlocks the full product set.

What is absent is the specification. No interface standard is named, so whether this runs on FHIR, HL7 version 2, a proprietary interface or file exchange is unknown. No marketplace or partner listing was located for either major system. Nothing states which direction data moves.

That last point is the entire question for one of these products. Sidekick Notes generates clinical documentation, and whether a generated note writes into the legal medical record automatically, lands in a draft queue awaiting clinician signature, or must be moved across by hand is the difference between a workflow product and a transcription tool. The same applies to assessment output: an outcome score entering the chart as a discrete field is materially different from a score displayed in a separate dashboard.

Ask which standard, which systems support write back, and whether generated notes require clinician signature before entering the record.

More than the bare cloud statement this axis usually receives, and short of a residency commitment.

What is published names the provider, the region scope and the availability model: Amazon Web Services, United States regions, multiple availability zone architecture. Networks are isolated in virtual private clouds. Customer data is segregated by unique identifier so that access to customer specific information remains with that customer, which describes logical isolation. Backups replicate across multiple availability zones, and business continuity and disaster recovery plans are stated with annual restore testing against recovery point and recovery time objectives.

What is missing is choice and commitment. Logical isolation is described but no dedicated or single tenant option is offered, and for an organisation whose competitors may sit in the same environment that is a procurement question rather than only a security one. No contractual residency commitment is stated, only a description of present practice. No customer managed encryption key option is mentioned, and no customer hosted or private deployment path exists.

One gap attaches to a named relationship. The company works with a federal medical technology consortium on service member suicide prevention, and federal health deployments typically require FedRAMP authorisation or a Department of Defense impact level. Neither is claimed anywhere, and nothing describes which environment serves that work.

Ask for the residency commitment in contract, the tenancy model, and the authorisation status covering federal work.

Cost is absent from every published surface. A dedicated pass found no pricing page, no unit of charge, no range, no implementation fee position, no minimum commitment and no pilot terms. Site navigation carries no pricing entry at all, and every path terminates at a demo booking form.

Two fragments are the whole of what exists. A professional association partnership offers members 20 percent off annual software subscription fees, which establishes an annual subscription basis and nothing further. TDScreen is free to any provider or patient, which prices one narrow tool and says nothing about the platform.

The product structure makes the silence more costly than usual. Seven separately named products span inquiry handling, intake, assessment, ambient documentation, group documentation, clinician coaching and life sciences research. Those carry entirely different cost drivers: documentation scales with clinician count, assessment scales with patient volume, inquiry handling scales with inbound demand. Nothing indicates whether they are licensed separately, bundled or metered, so a buyer cannot construct even a rough budget.

Nor is a return proxy supplied. The company argues 2 hours saved per clinician per day and fewer crisis escalations, which is a financial argument with none of its financial inputs attached.

Ask for the unit of charge per product, whether assessment is priced per patient or per facility, the implementation fee, and the minimum term.

Depth by clinical population is evidenced unusually well, depth by care setting is not evidenced at all, and that split is what holds this at B.

Population coverage is specific and backed by instruments rather than assertions. Depression, anxiety, post traumatic stress, postpartum depression and tardive dyskinesia each carry a named validated model, which makes the coverage claim checkable in a way that a list of served conditions is not. Postpartum screening and movement disorder screening are genuinely differentiated, and no other behavioral health record here addresses either. The named integration partners map to distinct settings, spanning addiction treatment, behavioral health and senior living record systems alongside the two major acute systems, and the life sciences line adds clinical trials and post market research as a separate buyer context.

What is missing is which products run where. More than 300 facilities are claimed and not one is characterised by level of care. Nothing separates residential from outpatient from intensive outpatient, which differ both clinically and commercially. Nothing states whether any model is validated for adolescent or paediatric populations, which represent a large share of behavioral health demand and use different instruments. Nothing addresses markets outside the United States, and acoustic and language models raise an obvious question about performance outside English that goes unanswered despite Spanish patient material being published.

Ask for the level of care breakdown behind the facility count, the paediatric position, and whether any instrument is validated in a second language.