Kennar Health

The company describes an AI workforce running agentic workflows, with a scribe capturing and structuring the encounter and an analysis layer surfacing quality gaps and risk opportunities from it. There is no system of record, services organisation or prior platform underneath, and every stated output is model produced. Graded on what the product is; note that the wider record is thin and a refresh should confirm this once more is published.

Agentic workflows operating at network scale with no described gate. The company states that providers can act on insights immediately within their existing workflow, which is the design goal, and nothing published describes what a provider is expected to verify before doing so. No confidence signal, routing threshold, abstention behaviour or required review step was located for either the scribe output or the surfaced risk and quality opportunities.

The compliance architecture discussed on the governance axis does supply an indirect control, since tying every code to a face to face encounter with an approved provider type constrains what can be submitted, but that is an eligibility rule rather than an accuracy check and should not be read as oversight of whether the code is right.

Numbers without method. Three figures are published, a 70 percent capture rate, a 0.2 risk adjustment factor lift and two hours saved per day, and none carries a denominator, a baseline, a population or a measurement approach. Capture rate in particular is meaningless without knowing capture of what, against what reference standard, and as judged by whom. No model or model family is named, no accuracy figure for coding or gap detection exists, and no evaluation methodology was located. The company does describe MEAT compliant documentation support, which at least names the criteria its output is meant to satisfy, and that is more specific than most of this category manages.

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 captured content informs model development. Basic questions about the capture path are also unanswered, including whether recordings are made at all as against real time processing, and how long anything is held.

One thing is named and it is worth crediting: the documentation criteria the output is meant to satisfy are identified explicitly, which is more specific than most of this category manages and tells a buyer what standard the coding support is written against. That is a standard, not a supplier.

The scope of what flows through the unnamed chain is wide, since the platform combines encounter capture with quality measure and risk adjustment analysis, so it holds both conversation derived content and the coded conclusions drawn from it, and the second category is what an audit would later examine. Ask what audio is captured and retained, whether encounters inform model development, and for a sub processor list covering both the capture and the analysis paths.

Nothing published establishes benefit or deployment. No customer is named, no funding is disclosed, no case study exists and no independent evaluation was located, and the company's public footprint is minimal. The three quantified claims are vendor asserted with no supporting material. Graded C rather than Not Rated because outcomes are claimed; nothing supports them.

Converted from Not Rated after a second search. No retention period, training use statement or de identification posture was located, and the scribe questions the prior note raised remain unanswered: whether recordings are made at all, how long anything is held, and whether captured content informs model improvement.

What the second pass established changes the frame around those questions rather than answering them. The company describes a scribe that listens to patient interactions and structures clinical data in real time, and in the same breath describes analysis that surfaces quality measure gaps and hierarchical condition category risk opportunities, with improved quality scores and increased revenue named as outcomes. Its own published questions include how customer data holds up under risk adjustment data validation audit.

That combination creates a stewardship question this index has not had to ask of a scribe before. A documentation tool that surfaces coding opportunities during the encounter is not summarising what happened afterwards; it is shaping what gets recorded while it is being recorded, and what gets recorded determines payment. The clinical record is therefore being optimised for reimbursement at the moment of its creation rather than mined for codes after the fact. That is lawful, it is what the customer is buying, and it means the record's provenance carries a commercial objective that a reader years later cannot see.

Ask what audio is captured and retained, whether encounters inform model development, and whether the coding prompt and the clinical note are generated by the same model against the same objective.

HIPAA compliance is claimed for the platform, with no business associate agreement terms published, which is the standard middle rung. One structural point worth establishing given the multi tenant architecture and the network buyers named: where an independent practice association or accountable care organisation contracts on behalf of participating practices, confirm which entity holds the agreement and how data is separated between participants who may be competitors.

Converted from Not Rated after a second search. No SOC 2, HITRUST, ISO 27001 or other attestation was located, and no trust centre or security page was found.

The prior note advised requesting evidence of tenant isolation specifically rather than a general attestation. The second pass confirms that is the right request, because the company states the architecture itself: it describes deploying across large provider networks through a multi tenant architecture supporting health systems, independent practice associations, accountable care organisations and managed services organisations. Multi tenancy is stated as a feature, which is honest, and it means the isolation question is not hypothetical.

It also identifies who the tenants are, and that is the part worth pausing on. Independent practice associations, accountable care organisations and managed services organisations in the same geography frequently compete for the same physicians and the same attributed patients. A platform holding encounter level clinical data and coding opportunity analysis for several of them at once holds competitively sensitive information, not merely confidential information, and the separation question is commercial as well as regulatory.

One segment observation belongs here. Ambient documentation is a category where an independently examined security posture has become the procurement baseline, and several competitors publish one. Measured against that norm the absence is conspicuous.

Ask whether isolation is logical or physical, whether tenants have separate encryption keys, what an administrator can reach across tenants, and whether any attestation exists.

No FDA clearance or device authorisation was located and none is expected, since the product captures documentation and surfaces coding and quality opportunities rather than diagnosing or recommending treatment.

Graded B rather than C because the regime that does govern is CMS risk adjustment rules and audit, and the company addresses that directly rather than leaving a buyer to discover it. Naming the regulator you actually answer to is the substitute for a clearance in categories that have no device pathway, and most vendors in this position say nothing at all. The detail sits on the governance axis.

Held off A because no external accreditation, audit result or third party examination of the coding output was located, which is what would move this from a stated position to a verified one.

The credit here is real and belongs first, because it is the sharpest compliance framing found anywhere in this index.

Asked how its data survives a RADV audit, the company gives a mechanism rather than a reassurance. It captures HCC codes prospectively at face to face encounters with MEAT compliant documentation support, states plainly that retrospective chart review is the primary driver of RADV clawback risk, and argues that a point of care model removes that exposure by design. It adds that every code traces to a face to face encounter with an approved provider type, meeting CMS encounter level evidentiary standards, and that output is standardised audit ready supplemental data rather than manual submission. That distinction, prospective capture versus retrospective review, is the single most useful thing a buyer in this market can understand about audit risk, and no other vendor in this index names it.

Held at C for three reasons. The company simultaneously markets a specific risk score outcome, a 0.2 RAF lift, which is the coding gradient stated as a target rather than a consequence. Eliminating exposure by design is an absolute claim and no audit outcome evidence supports it. And no fairness, subgroup or demographic performance disclosure of any kind was located, which matters where the product also drives quality gap closure.

Two passes located no accuracy figure for coding or gap detection, no error rate, no published limitations, no evaluation methodology and no warranty, indemnity or remediation commitment. Three figures are published, a capture rate, a risk adjustment factor lift and hours saved per day, and none carries a denominator, baseline, population or measurement approach.

Capture rate is the emptiest of the three, because capture of what, against what reference standard and as judged by whom are all unstated, and without them the number can describe either better documentation or more documentation. The substantive finding here concerns the record itself rather than the accuracy of any one output, and it is the first time this backfill has had to raise it.

The scribe listens during the encounter and, in the same flow, surfaces quality measure gaps and risk adjustment opportunities, with improved scores and increased revenue named as the outcomes. A tool that surfaces coding opportunities during the encounter is not summarising afterwards what happened, it is shaping what gets recorded while it is being recorded, and what gets recorded determines payment.

That is lawful and it is what the customer is buying, and it means the clinical record now carries a commercial objective at the moment of its creation that a clinician reading it years later cannot see. Ask whether the coding prompt and the clinical note are produced by the same model against the same objective, and what the vendor commits to when a surfaced condition is not supported.

EHR agnostic integration is claimed and a multi tenant architecture is described as deploying across large provider networks, which is the right shape for the buyer types named. Nothing is evidenced: no EHR vendor, no integration mechanism, no standard, no marketplace listing and no customer deployment was located anywhere, so an organisation cannot establish whether its own systems are supported. The supplemental data output is described as standardised and audit ready, which implies a defined submission format, and that is the one concrete interoperability claim on the record.

A multi tenant architecture is stated, which establishes hosted shared infrastructure and rules out a customer hosted option. Nothing further is published: no cloud provider, region, residency commitment or tenant isolation detail. Graded C because the delivery model is clear and everything else is absent.

No price, tier or pricing mechanism was located.

Given that the product is marketed on a specific risk adjustment factor lift, the contingent pricing question is the first one to ask. Establish whether any component of the fee varies with risk score movement, capture rate or quality bonus, since that would align vendor revenue directly with coding intensity. A plan or group buying on that basis is paying more the more diagnoses are captured, which is the same structure the index flags wherever it appears, and nothing published says whether it applies here.

Coverage is described by organisation type rather than by care setting or specialty. Four buyer types are named, health systems, independent practice associations, accountable care organisations and management services organisations, which is a coherent value based care footprint, and transitions of care is the one named clinical workflow beyond the encounter itself. No specialty, no care setting and no instrument level behaviour was located, and the ambulatory point of care focus is implied rather than stated.