Supanote

Ambient capture and note generation are the entire product: transcription, note drafting across the behavioral health formats, treatment plans, intervention planning and discharge notes, plus a personalization engine that adapts to a therapist's voice and a client workspace. There is no record system, billing arm or services layer underneath, and the plans are tiered by note volume and model capability rather than by anything the models do not touch. Remove the models and nothing is left to sell.

The oversight structure is described. Supanote drafts the note, the trust center states human review and approval before it is used, and the note reaches the practice's record system only when the clinician triggers Super Fill, so nothing files itself. The clinician sets tone, format and preferred interventions once, and the personalization engine follows that.

The client consent form Supanote supplies states plainly that the therapist reviews and modifies the note using their clinical expertise. What is missing is what happens after a draft is wrong: no acceptance rate, edit burden figure or confidence threshold is published, and nothing describes how errors are reported or corrected.

Supanote describes what its system does and names nothing behind it. It states that its models were trained on thousands of real therapy notes to judge the right level of detail, that a personalization engine learns a therapist's tone, format and preferred interventions, and that the top tier runs its most sophisticated model.

No model, provider, version or evaluation is published, no dataset description accompanies the training claim, and nothing describes how a model change is tested or announced. Stating what a model was trained on is not the same as stating how well it works.

The corpus is disclosed and the provenance is not. Supanote states that its models were trained on thousands of real therapy notes, which is candid and raises the question rather than answering it: whose notes, obtained on what consent basis. Its trust center answers the forward-looking half, stating that its models are never built from customers' sessions.

On enumeration there is less: the trust center carries a subprocessor section, but no model provider, model family or host is named in its published material, and nothing states which service performs transcription or note generation or what it retains while doing so. What reaches the chain is bounded on one side, since audio is deleted immediately after processing and identifiers are stripped from transcripts first.

Adoption and testimonials are what is on offer, with thousands of therapists claimed and a reviews page on Supanote's own site. Scale of use is not evidence of benefit. No study, controlled evaluation, accuracy benchmark or independent assessment is cited anywhere on its site, and its claims about note quality carry no method or denominator. Much of the comparative material about Supanote online is published either by Supanote or by competitors selling against it, so neither side is neutral.

The trust center answers retention, deidentification and training use directly. Recordings are processed and then deleted immediately, and a clinician can delete notes, transcripts or the whole session history, with automatic note deletion on the higher plan.

Personal identifiers are stripped from transcripts automatically before the note is produced, and the client consent form Supanote supplies tells the client that Supanote holds depersonalized transcripts with names, emails and other identifying details removed. On training, Supanote states that its AI models are never built from customers' sessions, and that its own team cannot access client notes, with role based access behind that.

One question sits underneath and is not answered: Supanote markets models trained on thousands of real therapy notes, and nothing states whose notes, obtained on what consent basis. Nothing is published on safety engineering either; the consent form acknowledges the system may carry unknown bias and leans on the therapist's review to catch it.

The Business Associate Agreement is published in full, entered into by Ionian Ventures Pvt Ltd trading as Supanote, and accepted in product. It requires subcontractors to agree in writing to the same or stricter restrictions on protected health information, and on termination Supanote must return or destroy it, continuing to protect anything that cannot be returned. Its liability under the agreement is capped at $10,000 in the aggregate.

The pricing page lists consent forms and the agreement as included from the entry paid plan, with a custom agreement for group practices; whether the 14 day trial is covered is not stated. Supanote also supplies a client consent form for therapists to use and states compliance with HIPAA, PHIPA, PIPEDA and GDPR.

Supanote runs a trust center at security.supanote.ai and presents the certifications as its own: SOC 2 and ISO 27001 are listed as complete, alongside HIPAA, PIPEDA and GDPR programs. What is not shown is the evidence behind them. No report, audit period, auditor or scope statement is published, and documents are not downloadable from the page, so neither the report nor its period can be checked from the site.

The posture page describes malware protection, device and container encryption, a personal data breach notification procedure, role based access, and an audit trail. Supanote's own published guidance for therapists makes the right point: no body awards HIPAA compliant status, and therapists are told to look for SOC 2, HITRUST or ISO 27001 with subprocessor disclosure and written retention policies.

The product makes no medical device claim and states no position on device status. It markets documentation, and the outputs that lean clinical are drafts the clinician signs: treatment plans, intervention planning and discharge notes, produced from a session the clinician conducted or dictated.

The rules that govern the product are payer documentation requirements, state licensure, recording consent law, and 42 CFR Part 2 where substance use treatment is documented, on which no position is stated. Supanote engages with consent as a product matter: it supplies a client consent form that names the technology, states that recordings are not stored and transcripts are depersonalized, lists risks including possible bias, and takes the client's signature, and it supplies the business associate agreement alongside it from the entry paid plan.

One gap follows from its own breadth: it claims compliance across HIPAA, PHIPA, PIPEDA and GDPR, and recording consent rules differ by state and province, so a single consent form cannot satisfy all of them.

There is no account of how Supanote's models are tested. No fairness statement, subgroup analysis, accuracy breakdown by accent, dialect or language, and no evaluation methodology appears on its site or trust center. What exists is policy language aimed at the client rather than the buyer: the consent form Supanote supplies states that the system may contain unknown bias in how it summarizes a session, and that the therapist's review is the mitigation.

Disclosing the risk to the person affected is a real step; it is not an evaluation. Two gaps are specific here. The product documents couples, family and group sessions, where an attribution error records one person's disclosure as another's, and no multi speaker performance evidence is published. And compliance is claimed across four regimes in three jurisdictions, implying users speaking more than one language, with no language coverage list or cross language performance.

Mechanisms for checking an output are published, with no commitment standing behind it. The trust center states human review and approval before a note is used, a secure audit trail, and user controlled deletion and sharing, and the note reaches the record system only when the clinician triggers Super Fill. The client consent form Supanote supplies goes further, telling the client the system may carry unknown bias and that the therapist reviews and modifies the note.

Nothing stands behind the note itself: no accuracy or error figure, no published limitations, and no warranty, indemnity or remediation commitment; the business associate agreement caps Supanote's liability at $10,000. Nothing describes how the system treats disclosed risk content, which is where an error carries the heaviest consequences.

Narrow by design and well aimed. Supanote's Super Fill writes a finished note into SimplePractice and TherapyNotes in one click, two systems that between them cover a large share of independent behavioral health practice, and the integration sits on the Pro plan and above. Input paths are flexible: live recording, dictation or uploaded audio. That is one directional writing into named systems, and it is where the exchange stops.

No other record system is named, nothing beyond note autofill is described, and no e-prescribing, laboratory, health information exchange or FHIR connection appears. A clinician on any other system moves the note by hand.

This is a hosted service used from a browser, with recording, dictation and file upload as input paths. Where it runs is not stated: no host, no region and no residency option appears on its site or trust center, and its privacy policy says only that data is transferred to the facilities and servers of its hosting providers and third-party service providers.

The retention posture shrinks what residency would protect, since recordings are deleted immediately after processing and finished notes are pushed into the practice's own record system, so exposure concentrates in the processing window, which is the part left undescribed. Residency is not academic here.

Supanote claims compliance with PHIPA, PIPEDA and GDPR, which carry expectations about where personal health information sits, and the contracting entity on its business associate agreement is Ionian Ventures Pvt Ltd. Its trust center also states that even its own team cannot access client notes.

One rate card with figures and a self serve start is published: a 14 day full trial at no cost, Starter at $19.99 a month for 40 notes with custom templates, treatment plans, couples sessions, consent forms and the business associate agreement, Pro at $39.99 for unlimited notes adding EHR integration, personalization, a client workspace and automatic note deletion, and XL at $69.99 adding group therapy sessions.

Group practices are quoted on request, with team permissions and a custom agreement. Monthly and yearly billing are offered, with the figures shown as starting rates, and the billing term each price reflects is not stated. The unit of charge is the note, which matches how a solo caseload scales.

Coverage is stated clearly and deeply within one specialty: the full behavioral health note format set including SOAP, DAP, GIRP, BIRP, SIRP, PIR and PIE plus custom templates, more than 20 therapy modalities including CBT, DBT, EMDR and internal family systems, and individual, couples and group sessions, with group work sold on the top tier. Supanote names psychiatrists, psychologists, social workers and coaches among its users. Nothing published validates the coverage by setting: no study, named deployment or outcome is tied to any modality or session type, including the multi speaker work that is hardest to document.