Nudge AI

Capture, note generation, template conversion, coding and the session and patient level analytics are all model output, with no services layer underneath.

A review step is described, with the clinician reviewing and refining the draft through instruction based editing and able to regenerate against a different template.

What pushes this to C is what happens without review. Diagnostic codes are assigned automatically, and the billing assistant automatically generates add on justifications. Writing the justification for a billing code is a materially different act from suggesting the code, because it produces the documentary evidence a payer would examine rather than a recommendation a clinician evaluates.

Combined with one click population of structured fields, checklists and radio buttons, a clinician can move from recording to a fully coded and justified chart entry with very little friction, and no confidence threshold, accuracy figure or abstention behaviour is published for any of it.

No accuracy figure, model card, named models or evaluation methodology located. Positioning rests on superlatives, describing itself as the most advanced documentation platform built for healthcare professionals, and the EHR count is stated inconsistently across sources at more than 50 in some places and more than 140 in others.

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, and nothing states whether customer content is used to train or improve models, which is the weighted question in a segment whose input is recorded psychotherapy. Named encryption and audit logging protect and record access without saying who holds it.

The consent tracking feature credited on the liability axis is worth a note here too, because it creates a useful asymmetry a buyer should notice: this vendor can tell a patient that their consent was recorded, and cannot tell them which companies process what they said. Consent to be recorded is not consent to any particular chain, and a consent record that does not enumerate recipients records the agreement without describing what was agreed to. Ask for a sub processor list, an explicit training position, and whether the consent record names who receives the content.

No study, controlled evaluation, accuracy benchmark, third party rating, named customer or deployment count located. The single performance claim, saving more than eight hours weekly, carries no denominator, cohort or method.

Three named controls, one of which is close to unique in this index. AES-256 encryption and audit logging are stated, and so is consent tracking, which makes patient consent a recorded, auditable object rather than a policy or a suggested script. Across this whole wave only four other vendors address consent at all, and none ships it as a tracked feature.

Held at B because no retention schedule, de identification practice or statement on whether customer content trains models was located, and PHI stewardship is the weighted axis in this segment where the input is recorded psychotherapy.

HIPAA compliance is stated and business associate agreements are supported, but with a qualification worth noticing: BAA support is described as being for enterprise use.

A solo therapist on an individual subscription should confirm directly whether they receive one, because on this reading they may not, and a behavioral health clinician recording sessions without a BAA in place is exposed regardless of how the software behaves.

A second pass located no SOC 2 report of either type, no HITRUST certification, no ISO 27001 and no trust centre. Encryption and audit logging are described, and both are controls rather than independent examinations of controls.

The delivery mechanism creates a security question this axis would not otherwise reach, and it is the most important thing on this record. The product is delivered substantially through a browser extension that reads the record system screen and writes into its fields. An extension with permission to do that has access to whatever is rendered in that browser, which is not limited to the encounter being documented. It extends to any record the clinician opens in that session, and potentially to other systems accessed in the same browser.

That is a materially different access surface from a server to server integration, and it sits on the clinician's own device rather than inside a controlled environment. The questions a buyer should ask are specific: what page permissions the extension requests, whether its access is scoped to particular record system domains or granted broadly, what it transmits from pages other than the active encounter, and whether an organisation can restrict or audit it centrally rather than relying on each clinician's browser.

The data involved raises the stakes. This is a behavioural health product handling psychiatry, therapy and medication management sessions, with templates supporting substance use disorder programmes, so the material passing through that extension is among the most sensitive in this index.

The grade reflects what a counterparty can verify before contracting rather than a judgement that controls are absent. Ask for the report and its scope, and separately for the extension's permission manifest.

No clearance claimed and none required for documentation. No United States device pathway attaches. What governs is payment integrity, and this is the sharpest instance of it in this category.

The product generates diagnostic codes with supporting rationale, identifies procedural and add on codes with justifications, analyses session complexity to set the evaluation and management level, and flags what it calls missed revenue opportunities. The vendor quantifies the effect: a nearly twenty percent lift in relative value units per visit, a stated annual revenue recovery figure per clinician, and a finding that a fifth of medication management sessions had missed billable add on codes which the product then identified, naming the psychotherapy and crisis code families specifically.

Set that against the field literature. A 2025 policy analysis in a peer reviewed digital medicine journal documented ambient scribes producing an eleven percent rise in work relative value units and a fourteen percent increase in documented risk adjusted diagnoses per encounter, and treated those as a policy concern about coding intensity. This vendor markets roughly double that lift as a product benefit. The behaviour the literature describes as an open question is here the value proposition.

The structural point is sharper still, and it is the reason this record matters beyond its size. The system does not only propose a code. It generates the written justification supporting that code. Documentation is supposed to support the code because it records what happened; here the same system selects the code and authors the text that defends it, so the justification cannot function as an independent check on the selection. Ask what the justification is derived from, and whether it can ever fail to support the code the system chose.

The audit framing deserves scrutiny too. Notes are described as audit proof and insurance proof, and the product as shielding a practice from audit risk while maximising reimbursement. Those are opposite ends of a trade off that regulators treat as opposed, and no documentation product can deliver both by construction.

One further regime applies and is unaddressed: the product supports substance use disorder programme templates, and those records fall under a separate federal confidentiality regime with its own consent rules.

At the aggressive end of C, and the language is the vendor's own. It describes identifying billing opportunities such as psychotherapy add ons, naming CPT 90833 specifically, assessing session complexity to recommend medical decision making codes, helping clinicians maximise reimbursement, and facilitating increases in reimbursement per session. Generating the justification for an add on code is the sharpest form of this found anywhere in the wave, because the system is producing the very documentation that would be examined if the claim were questioned.

Real counterweights exist and belong in the same paragraph: the notes are framed as audit proof and as demonstrating medical necessity, consent is tracked, and actions are audit logged, all of which point at defensibility rather than only at revenue.

What is absent is any fairness statement, subgroup analysis or performance disclosure, and no language coverage beyond English.

One shipped feature carries this grade and it is close to unique in this backfill. Consent tracking is a named control, which makes patient consent a recorded and auditable object rather than a policy paragraph or a suggested script for the clinician to remember. That matters on this axis specifically.

Across the whole ambient wave only a handful of vendors address consent at all, most disclaim it entirely and place the burden on the clinician, and none of the others ships it as something the system records. Recourse begins with the affected person having been asked, and an auditable record of asking is what makes that checkable a year later when someone disputes whether a session should have been captured. Audit logging and named encryption sit alongside it.

Held at C because nothing stands behind the output and the published claims work against confidence. No accuracy figure, error rate or evaluation methodology was located, positioning rests on a superlative describing the product as the most advanced documentation platform for healthcare professionals, and the record system integration count is stated inconsistently across sources at more than 50 in some places and more than 140 in others. Inconsistent figures on a vendor's own material are a poor basis for reliance. Ask what the consent record captures and who can produce it later, and for an accuracy figure with a definition.

A meaningfully better transfer mechanism than the paste workflows typical at this price point. The Chrome extension operates from the record screen and populates fields with one click including checklists, radio buttons and structured fields, not just a block of narrative, which is the difference between filling a form and pasting into one. Breadth is claimed across 50 or more EHRs in most materials and 140 or more in another, an inconsistency worth resolving.

Held at B because this is still browser level field population rather than a sanctioned data channel, so it inherits the brittleness and audit trail questions this index notes for robotic process automation.

Confirmed and materially clarified. No hosting region, no residency option and no subprocessor list were located, and nothing establishes whether a third party model service processes the session or what it retains.

What the second pass did establish is the integration mechanism, and it reframes the vendor's headline compatibility claim. The product reaches the record system through a browser extension described as working directly from the record system screen and filling every field with a single click. Compatibility is advertised across a very large number of record systems, with the figure varying between vendor and third party sources.

That breadth is evidence about the mechanism rather than about engineering depth. Supporting well over a hundred record systems is achievable through a browser extension precisely because the extension operates on the rendered page rather than through each system's own interface. So the compatibility claim and the absence of permissioned integrations are the same fact stated two ways, and a buyer should read it that way rather than as unusually broad interoperability work.

This is the third vendor in this lane reaching the chart by manipulating what a user sees rather than through an interface built for the purpose, alongside a robotic process automation approach and a browser page manipulation approach elsewhere in the category. The questions are the same in each case. Establish whose credentials the automation runs under, how its writes appear in the record system's audit log, whether the permission model constraining a human user constrains it equally, and what happens to the integration when the host system changes its interface.

Ask also for the hosting region, whether it can be pinned by contract, and which model provider processes the session.

A genuine free trial is published at no cost with five notes and no credit card required, which lets a clinician evaluate the output before committing. Beyond that, subscription pricing of roughly 85 to 99 US dollars a month appears on a third party directory rather than on the vendor's own materials, and enterprise pricing is not described at all despite the platform offering member management and analytics for organisations.

Behavioral health across the clinician types that actually practise it, psychiatrists, psychologists, therapists, counsellors and psychiatric mental health nurse practitioners, producing psychiatric evaluations, therapy progress notes and treatment plans.

Its distinguishing depth is by LEVEL OF CARE rather than by note format: it ships templates for partial hospitalisation daily progress notes in substance use disorder programmes written to Medicare and insurance documentation requirements, and for day treatment settings. Those are documentation regimes with their own regulatory expectations and almost nothing else in this index addresses them. Templates are created by pasting an existing note for the system to replicate. English only.