Reteta

Ambient capture, note generation, pre charting, after visit summaries, referral letters and the coding output in Bcoder are all model produced. No human scribe tier and no services layer. The company is a product company launched out of a venture studio rather than an existing services business adding AI, and both products are sold as the thing itself rather than as a feature of a platform.

The review gate is described plainly, with clinicians reviewing and signing generated notes, and the explainability design is a real oversight mechanism rather than a slogan: if a clinician can trace any sentence in a note back to the moment in the conversation it came from, verification becomes an act they can actually perform in seconds rather than a re reading of the whole note against memory. That is the same property that makes Abridge the auditability benchmark in this lane.

Held at B rather than A because the claim is asserted without a demonstration or an error rate, because coding suggestions and CDI prompts surface in real time during the encounter without any published confidence threshold or abstention behaviour, and because the figure that 80 percent of notes are accepted without edits is presented as a benefit when it also describes how often the review step changes nothing.

Graded B for one disclosure this index has been asking the whole category to make, and held at B rather than higher because the numbers around it do not reconcile.

Reteta reports 86 percent accuracy across diverse accents and dialects. No other ambient vendor found in this wave publishes an accuracy figure broken out by accent, and it is the exact ask this index has put to Augnito, whose entire differentiator is accent coverage but which converts it into unfalsifiable marketing. Reteta also describes explainability as an architectural property, stating every note entry traces back to its point in the conversation.

Against that, the company publishes at least four accuracy claims that cannot all describe the same thing: 86 percent across accents, up to 95 percent accuracy on the homepage, 80 percent of notes accepted without edits, and 90 percent coding accuracy for Bcoder, with no methodology, denominator or evaluation design for any of them. Up to 95 percent is the unfalsifiable shape this index flags. The accent figure also sits in a blog post comparing Reteta with a named competitor, which is marketing content rather than technical documentation.

Ask for the accent breakdown as a proper evaluation and this moves to A.

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. End to end encryption is stated for protected health information in transit and at rest, which protects content without naming who processes it. One capability creates a content type worth asking about separately, because it is derived rather than captured.

The after visit summary generation reads patient language preference and infers health literacy level in order to adapt what it produces. An inferred health literacy level is an attribute the system assigns to a person, it did not come from the patient and the patient is unlikely to know it exists, and it is the kind of derived characteristic that can persist and travel further than the summary it was created for.

Establish which patient attributes the system derives, where they are stored, whether they are written to the record, and whether they are excluded from any model improvement process. Ask for a sub processor list alongside that, and for whether any third party model provider processes summary generation.

No study, controlled evaluation, third party assessment or named customer located. Adoption is described only as hundreds of satisfied healthcare providers, with no organisation named and no deployment count. Every performance figure is vendor stated without method: two hours saved per clinician per day, SOAP notes in under two minutes, burnout reduced by up to 75 percent, an 85 percent increase in claim approval, 70 percent coding productivity gain.

Presence on the Elion and AVIA marketplaces is worth noting but not crediting as validation: the Elion profile is explicitly marked unclaimed, meaning it was generated by the directory rather than submitted and verified by the vendor.

End to end encryption is stated for PHI, and the company describes safeguarding data in transit and at rest. Nothing further was located: no audio retention schedule, no deletion commitment, no de identification practice and no statement on whether customer content is used to train models. Encryption without a retention answer is the ordinary position in this lane rather than a strong one.

Given that the after visit summary generation reads patient language preference and infers health literacy level, a buyer should also ask what patient attributes the system derives and stores in order to do that.

At the upper end of B, on a specific and checkable claim rather than a generic one. Reteta states it is HIPAA One compliant, naming an actual third party HIPAA compliance certification programme rather than simply asserting that the product is HIPAA compliant. Naming the assessing body is meaningfully better than the bare claim most of this lane makes, and it gives a buyer something to request evidence of.

Held at B rather than A because no business associate agreement terms, availability or tier gating are published, and because the certification date and scope were not located. Doximity holds the A on this axis because coverage there is automatic and universal with no separate step.

A second pass located no SOC 2 report of either type, no ISO 27001, no penetration testing statement, no audit report and no trust centre. The vendor describes end to end data protection, which is a claim about design rather than an examination of it.

One attribution from the earlier assessment stands and is worth restating, because it is the kind of item that inflates a compliance list. The certification this vendor holds is a health privacy rule compliance attestation, which is a different object from a security controls attestation. It speaks to a documented compliance programme rather than to an independent examination of technical and organisational controls, and it is graded on the health privacy axis rather than here. This index applies the same distinction throughout: laws, regulators, conformity marks and control attestations are four different kinds of thing, and a list mixing them reads as more than it contains.

The absence matters in proportion to what the platform holds and who it sells to. This vendor integrates with the major hospital record systems through interface standards and writes notes, codes and summaries back into them, so it holds encounter audio, generated documentation and coding output for organisations whose procurement functions expect an examined posture as a matter of routine. Selling into that buyer without one is a harder position than selling to a solo practice without one.

The grade reflects what a counterparty can verify before contracting rather than a judgement that controls are absent. Ask for the report, its type and period, and for whatever external testing has been performed.

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

What governs is payment integrity, and this record sits at the sharp end of it. The platform pairs an ambient scribe with a dedicated coding product delivering real time coding suggestions. The vendor markets real time coding visibility as helping clinicians maximise reimbursement accuracy, describes compressing the clinical documentation improvement cycle from days to minutes by identifying documentation gaps during the encounter, and claims a large increase in claim approval.

The field has now published on precisely this, which converts the concern from an index opinion into a documented pattern. A 2025 policy analysis in a peer reviewed digital medicine journal describes the business case for ambient scribes shifting from clinician wellbeing toward revenue capture through more intensive coding, and reports measured effects: an eleven percent rise in physician work relative value units and a fourteen percent increase in documented hierarchical condition category diagnoses per encounter at one health system, more high level evaluation and management visits billed at another, and an oncology study finding documented diagnoses per encounter rising from three to four point one.

That analysis names the question to put to this vendor: whether better coding reflects better documentation of care actually delivered, or more intensive coding of the same care. Identifying gaps during the encounter rather than afterwards is the sharper version, because it shapes what the clinician documents while they are documenting it.

One thing is credited. The vendor describes its scribe as explainable and its coding as traceable to the encounter. A suggestion a clinician can trace back to what was said is reviewable in a way an unexplained one is not, and that is the right direction for a product on this gradient.

Genuinely mixed, and both halves matter. The positives are real and unusual. Publishing an accuracy figure broken out by accent and dialect is the single most requested disclosure in this index's speech coverage and almost nobody does it. Generating the after visit summary to the patient's own health literacy level and preferred language is an accessibility design decision rather than a marketing line, and it addresses a failure mode the category otherwise ignores entirely.

Explainability gives a clinician a way to challenge output. Against that, the coding posture puts it on the gradient: CDI capabilities identify documentation gaps UPFRONT and are described as compressing the CDI cycle from days to minutes, which is the point of care prompting pattern this index flags in Cleo and Solventum, and the outcome sold is an 85 percent increase in claim approval alongside claim ready notes.

That framing sits at the defensive end, denial avoidance rather than coding intensity, which is nearer ScribeEMR and LucasAI than MarianaAI. No fairness statement, subgroup analysis or model card accompanies the accent figure.

This record is a hair short of the band above it, and saying exactly what would close the gap is more useful than the grade. The vendor publishes an accuracy figure broken out by accent and dialect, at 86 percent. No other ambient vendor located in this wave publishes accuracy by subgroup at all, and it is the precise disclosure this index has been asking the category for, because speech recognition error falls unevenly across accents and that unevenness lands on patients rather than on the vendor.

A vendor willing to publish a subgroup figure that is not flattering is doing something the rest of the lane avoids. Traceability supports it: every note entry is stated to trace back to its point in the conversation, which is a real correction mechanism at the moment of review. What holds it here is that the numbers do not reconcile.

At least four accuracy claims are published that cannot all describe the same thing: the accent figure, a homepage figure using the unfalsifiable up to construction, a proportion of notes accepted without edits, and a separate coding accuracy figure, with no methodology, denominator or evaluation design behind any of them. The accent figure also sits in a blog post comparing the product with a named competitor, which is marketing rather than technical documentation. Publish the accent breakdown as a proper evaluation with a stated method and sample, and reconcile the four figures, and this becomes one of the strongest records in the lane.

Described as standards based and bidirectional: secure APIs over HL7 and FHIR, writing SOAP notes, coding and summaries back into the record, with Epic, Cerner, athenahealth and eClinicalWorks named, plus access to patient information and appointments from inside the platform. Naming a standard and naming systems is the right shape and puts this above the transfer tier.

Two things hold it at B. The phrase bolts onto any EHR is the universality claim this index treats as a warning sign, since in this lane universal compatibility is usually a consequence of not integrating rather than of integrating well, and the named list ends in and more without a supported systems page. And no customer is named on any specific system, so the integrations are described rather than demonstrated.

Establish in a sandbox whether write back lands as discrete coded fields or as a text blob.

Confirmed rather than changed. Delivery is a cloud platform. No hosting region, no residency option, no subprocessor list and no on premise path were located.

What is better established is the integration surface, and it deserves crediting because two other vendors in this lane reach the chart by considerably less controlled means. This product connects through secure interfaces and the standard health interoperability protocols, writing notes, codes and summaries back into the record system that way. That is a permissioned integration rather than a browser manipulation or robotic process automation approach, so its actions carry their own identity and the record system's own access controls apply to them.

What remains unanswered is the question this axis asks first of any ambient scribe. 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. The coding product raises the same question separately, since a real time coding suggestion produced during an encounter is generated somewhere and from something.

Named integrations include the largest hospital record systems in the United States. A buyer of that size will have a residency position of its own and a subprocessor review process, so the gap here is likely to be closed in contracting rather than in published material. That does not help a buyer trying to compare vendors before entering a sales conversation, which is what this axis measures.

Ask for the hosting region, whether it can be pinned by contract, the subprocessor list, and which model provider processes the encounter and the coding.

No price, unit of pricing, tier structure or trial published, and no third party estimate located. The routes in are a demo request and marketplace listings, neither of which exposes a rate.

More than 15 specialties with specialty specific templates, spanning ambulatory practice and hospital or health system settings. The output range is wider than most small vendors attempt: SOAP notes, pre charting before the visit, after visit summaries, and referral letters to other specialties, which is a document type many scribes never produce.

The distinguishing element is the after visit summary, generated with education materials, treatment regimens and caregiver notes tailored to the patient's health literacy and preferred language. That reaches the caregiver as a reader, not just the clinician and the payer. Held at B because the specialty count is modest against the leaders in this lane and no specialty instrument is named, which is this index's test for real domain depth as opposed to a template library.