TheraPulse

Transcription and note generation are the entire product, with a secondary research lookup tool alongside. No services layer or adjacent business underneath.

Conventional draft and review, with a genuinely useful alternative path. Clinicians who prefer not to record can use a dictation route where they speak their own reflections and key points after the session and the system structures those into a note, so the clinical content originates with the clinician rather than from captured conversation. Offering a non recording path at all is uncommon and it materially changes both the consent and the accuracy question. Held at B because no acceptance rate, edit burden figure or confidence threshold is published.

No accuracy figure, model card, named models or evaluation methodology located. The performance claims are speed and time saved rather than correctness: notes in around 60 seconds and up to two hours saved daily.

Nothing identifies any party in the chain: no model or model family, no foundation model provider, no hosting arrangement and no sub processor list was located in two passes. The boundary is partly answered and the part that is answered is the strongest of its kind in this segment: recordings are used only for processing and deleted immediately after transcription, which removes the most sensitive artefact rather than managing it. Two limits keep this at the floor.

The commitment is drawn around one artefact, covering recordings and saying nothing about transcripts or generated notes, and those are the items that persist and carry the clinical substance. And nothing states whether either is used to train or improve models, which is the commitment that separates the top grades in this segment. One published word is doing unclear work and should not be read as a control.

Session data is described as anonymised, with no method, scope or pipeline position given, and a psychotherapy note is inherently identifying in ways structured field removal does not address, since the content itself describes a specific person's circumstances. Ask what anonymised means here, what happens to transcripts and notes, and for a sub processor list.

No study, controlled evaluation, accuracy benchmark, third party rating, named customer or documented deployment located. Claims are vendor generated and the reviews available are affiliate style write ups rather than assessments.

The strongest deletion commitment located in the behavioral segment, on the most sensitive audio in this index. TheraPulse states that recordings are used only for processing and are deleted immediately after transcription, alongside AES 256 encryption and a claim that session data is anonymised. Immediate deletion of psychotherapy audio removes an entire category of exposure rather than managing it.

Three things hold it at B rather than A. Anonymised is doing unclear work, since therapy notes are inherently identifying. The commitment covers recordings and says nothing about transcripts or generated notes. And no statement was located on whether either is used to train models, which is the commitment that earns Mentalyc an A in this same segment.

HIPAA compliance stated consistently. Business associate agreement terms are not published for inspection, and the product is self serve so a clinician can begin before any agreement conversation.

A second pass located no SOC 2 report of either type, no HITRUST certification, no ISO 27001 and no trust centre. What is published is a cipher and a compliance claim: AES 256 encryption, described on the vendor's own site as the same encryption banks use, alongside a statement of health privacy compliance. A cipher is a control rather than an attestation, and the banking comparison is marketing language rather than assurance.

The usual segment norm argument needs adjusting here, and adjusting it makes the finding sharper rather than softer. This product is sold to individual therapists and small practices at monthly tiers metered by session volume. That buyer will never run a vendor security review, has no procurement function, and is not in a position to read a service organisation control report even if one existed. So the observation that competitors publish attestations carries less weight than it does for an enterprise product.

What replaces it is a harder point. The buyer is the least equipped in this category to assess a vendor, the data is among the most sensitive in the index, being recorded psychotherapy sessions, and the regulatory duty does not move: the therapist remains the covered entity and carries the obligation to satisfy themselves about a business associate. A vendor selling into that asymmetry has a stronger reason to publish independent assurance than an enterprise vendor does, not a weaker one.

Fairness is still due to the stage of the company. A full attestation programme is a material cost at this size, and this grade reflects what a counterparty can verify rather than a judgement that controls are absent. Ask for the business associate agreement and for whatever external testing has been done.

No clearance claimed and none required for documentation. No United States device pathway attaches to a note the clinician reviews and signs.

What governs behavioural health documentation is not device regulation but a three part structure this index has set out elsewhere in the category: psychotherapy notes carry protections beyond ordinary clinical records, substance use disorder treatment records fall under a separate federal confidentiality regime with its own consent rules, and recording a therapy session engages state consent law which varies by state and attaches to the recording itself rather than to the note.

This record is graded below a peer that engages with that structure, and the comparison is instructive because it turns on one document. A competitor in this same segment publishes client consent language for clinicians to use, including an explicit statement that the client may decline at any time without it affecting their care. This vendor's published guidance on the same conversation recommends framing it around the benefits to the client, such as receiving detailed summaries for reflection and engagement.

That advice is not improper and the benefits described are real. But a framing built around persuading a client to agree is a different instrument from language designed to secure informed consent, and in a therapeutic relationship the difference matters more than it would elsewhere, because the power asymmetry between clinician and client is part of what consent rules exist to address. A vendor supplying adoption guidance rather than consent wording has left the harder half to its users.

On scope, the product also generates summaries and insights described as supporting treatment planning and decision making, which is the decision support boundary rather than documentation.

Consent is addressed directly, which almost no vendor in this index does, and it should be read with both hands. On the credit side, TheraPulse states that recording is OPTIONAL and only done with CONSENT, and it backs that with a real non recording workflow rather than leaving refusal as a dead end. On the other side, its guidance coaches clinicians on how to FRAME the request, recommending the conversation be built around client benefits such as receiving session summaries.

Scripting a consent conversation to improve acceptance deserves scrutiny anywhere, and more so in psychotherapy, where the power differential is pronounced and the material is the most protected category of clinical data. Separately, no fairness statement, subgroup analysis or accent and dialect disclosure was located.

Two passes located no accuracy or error figure, no published limitations, no evaluation methodology and no warranty, indemnity or remediation commitment, and the performance claims measure speed and time saved rather than correctness.

The segment makes that heavier than it looks, because a psychotherapy progress note supports a clinical formulation, is submitted for reimbursement, and is discoverable in custody, criminal and civil proceedings, so an inaccurate note is read years later by parties the client never anticipated.

Risk content is the case that most needs a published position, since a note that omits or misstates disclosed suicidal ideation, abuse or substance use carries immediate clinical and legal consequences, and nothing published describes how the system treats it.

One control on this record is genuinely strong and belongs to the other axis rather than this one, and the distinction is worth drawing because a buyer will otherwise credit it here: recordings are deleted immediately after transcription, which removes an entire category of exposure and is the best deletion position in this segment. It governs what happens to the audio, not whether the note derived from it is correct, and the note is the artefact that persists and gets read. Ask for accuracy on risk content specifically, and for what the vendor commits to when a note is wrong.

Compatibility rather than integration. The vendor describes the product as working with most browser based electronic record and telehealth platforms, which in practice means the note is produced alongside the record system and moved by the clinician rather than written into it. For a solo therapist that is usually sufficient and it preserves a firm human gate, but no named integration, write back mechanism or structured field population was located.

The deletion posture is confirmed: recordings are used for processing only and destroyed immediately after transcription, and session data is described as anonymised. As with other minimal retention vendors, that materially reduces what residency would protect, because the persistent surface is small by design.

The exposure moves rather than disappearing, and it concentrates into the processing window, which is the part left undescribed. Nothing published establishes where transcription and note generation run, whether a third party model service performs either, which provider that would be, or what it retains. A commitment that the vendor keeps nothing says nothing about what a subprocessor keeps. No hosting region, no storage location for the notes themselves, and no subprocessor list were located.

Residency is not academic here. Compliance is claimed against the Ontario health privacy regime alongside the United States rule, and Ontario carries explicit expectations about where personal health information sits. A vendor claiming that regime without stating a hosting region has left its Canadian buyers unable to answer the first question their regulator asks.

One feature makes this record different from a general scribe and deserves its own question. The product markets multi voice detection for couples and group therapy. A group session recording contains the disclosures of several people at once, most of whom are not the subject of the note being generated. Where that audio is processed, what is retained about the non subject participants, and whose record their words end up in are questions no architecture diagram for a one to one scribe would raise. Ask them specifically before using this in group work.

Published and sensibly structured, tiered by session volume rather than by seat: reported at roughly 29 US dollars a month for 30 sessions, 49 for 50 and 99 for 100, with a fourth tier above. Volume pricing suits part time and building practices, and it is the right shape for a market where caseloads vary widely. Held at B because the figures were located through a third party review dated late 2024 rather than confirmed on the vendor's own pricing page in this pass, so they may be stale.

Behavioral health across a wide range of clinician types, naming therapists, psychologists, counsellors, clinical social workers, psychiatrists, marriage counsellors and group therapists, with speech pathology also referenced. Note formats cover SOAP, DAP and BIRP. Its multi voice detection extends coverage to couples, family and group sessions, which is a harder documentation problem than individual therapy and which most scribes do not attempt. Narrower than JotPsych, which adds psychiatric medication management depth, but broader than the therapy only tools on session type.