TransformativeMed

The asset is knowledge engineering and interface design, not a model. The company's own long standing description of its method is a proprietary knowledge engineering process that redesigns the record system to support each acute care team's workflow, and that is exactly what the product does: it decides what a cardiologist's patient list should show and in what order, then renders it inside the record using the vendor's native extension toolkit.

That is skilled clinical content work and it is not learning. The artificial intelligence in the current offering arrives largely through a partnership announced in September 2025 with Vantiq, whose real time platform supplies the agentic capability and holds the patented technology described in the announcement, positioning this company as the preferred agentic platform for Oracle Health clients rather than as the builder of the agents.

That makes this the fifth instance in this index of a workflow incumbent carrying a partner's intelligence. Reversible: if the company publishes what its own models do, or the agents built on the partnership turn out to be substantially its own work, this grade moves.

Nothing is decided or actioned automatically. The platform assembles information, routes tasks and alerts, and synchronises state across a care team, and clinicians do the rest. One configuration property is worth the same scrutiny applied to Azra in this index: the alerting module lets clinicians subscribe to the specific care alerts they want.

Subscription is a sensitivity control expressed as a preference, and a clinician choosing which alerts to receive is also choosing what not to be told. That is arguably the right way to handle alert fatigue, since the person carrying the pager decides, but no published material describes what a given subscription configuration omits, so the choice is made without any information about its cost.

The older half of the product is inherently inspectable and the newer claims are not. Specialty workflows built as configuration on a documented extension toolkit are knowable to the institution running them: a hospital can see which fields appear on which view and why, in a way no learned model in this index permits. That is the same structural advantage recorded against rules based surveillance platforms. What holds this at C rather than B is the labelling.

The company now describes embedded intelligence and artificial intelligence driven automation and positions itself as an agentic platform, without specifying what is automated, what decides, or which component belongs to the partner. TheraDoc earned a higher grade in this index precisely because it described a rules engine as a rules engine and made no artificial intelligence claim at all; VigiLanz was marked down for the phrase rule based artificial intelligence. This record sits between them.

Strong market validation, no clinical literature. The company was named 2026 Best in KLAS winner for clinician digital workflow, following earlier recognition in the analyst firm's emerging solutions work where it placed first for improving the physician experience and second for improving outcomes.

Deployments are long lived and verifiable in customer announcements: University of Tennessee Medical Center since 2013 and extended to 650 beds, University of Washington Medicine including Harborview, University of Alabama at Birmingham at 1,157 beds, a Veterans Affairs division, and King Faisal Specialist Hospital. A vendor measured figure of up to 45 minutes saved per provider per day is claimed.

Two limits keep this at C. Analyst awards measure customer satisfaction and product experience rather than clinical outcomes, and this index has consistently recorded such rankings as market signals with their own sample and scope conditions. And two retrieval passes located no peer reviewed publication, which is a gap a decade of academic medical centre deployments should have closed.

The architecture is favourable and the documentation is absent. Because the workspaces are embedded inside the customer's own record system using its native extension framework, most data handling happens where the record already lives, which limits how far information travels compared with a hosted service pulling extracts.

Two passes located no published statement on retention, on what the mobile application caches on a personal or shared device, or on how the partner's agentic layer receives and holds data. That last point is the newest and least examined: adding an external real time platform to a product that previously ran inside the record changes the data path, and nothing published describes the new one.

Two retrieval passes located no HIPAA statement, no Business Associate Agreement terms and no privacy or legal page. Graded on published posture. Agreements certainly exist, given deployments inside large academic health systems and a federal facility.

The mobile component deserves specific attention in any contract here, because an application that both reads patient data from and writes it back to the record on a clinician's own phone raises device management, loss and remote wipe questions that a purely desktop embedded product does not.

Two passes located no SOC 2, no HITRUST, no ISO 27001, no trust centre and no vulnerability disclosure policy. One detail makes this absence unusually likely to be correctable rather than genuine: the platform is deployed at a Veterans Affairs division, and federal deployment normally requires the product to clear a formal security authorisation process. If that assessment exists, it is a stronger attestation than most commercial certifications and it is not published anywhere located.

This is the same pattern noted against TheraDoc, which serves dozens of federal medical centres through a reseller and likewise publishes little about it. Vendors serving federal customers hold the best evidence in this market and mention it least.

No clearance, authorisation or submission located and none claimed or needed. Presenting record data in a reorganised view and routing tasks between clinicians is not device functionality, and the product sits comfortably outside device regulation.

The boundary worth watching is the same one now appearing across several records in this index: as the agentic layer takes on automation of clinical work rather than presentation of it, the question shifts from how information is displayed to what the software decides to do, and that is a different regulatory conversation. Nothing located suggests the current product has crossed it.

Nothing published, and there is a specific governance question that this product class raises and nobody in it addresses. Specialty specific workflow configuration is an editorial act: deciding what appears at the top of a nephrologist's patient list, which values are surfaced on rounding, and what counts as discharge readiness are clinical judgements encoded by the vendor's knowledge engineers on behalf of every institution running the product. Those judgements age.

Practice changes, formularies change, quality measures change, and a view that quietly stopped reflecting current practice keeps rendering with institutional authority behind it. This index recorded the absence of any drift monitoring concept across rules based clinical surveillance, and the same gap applies to interface design with less scrutiny, because a stale alert eventually annoys someone whereas a stale layout simply becomes invisible. No versioning policy, review cadence or subgroup consideration was located.

Depth of record system integration is not a feature of this product, it is the product, and it is genuinely deep. The workspaces are built with the record vendor's own extension framework and run inside the record rather than beside it, so clinicians do not change application, log in twice or reconcile two views of the same patient, which is the failure mode that kills most overlay products. Data flows both ways, including from the mobile companion back into the record.

Task synchronisation, alert routing and secure messaging operate across the whole care team inside that same environment. It runs on Oracle Health Millennium and on Epic, at large academic medical centres, a federal facility and internationally, and the analyst award it won in 2026 is specifically for clinician digital workflow, which is the category this axis measures.

Running inside the customer's own record instance through the native extension framework is the favourable answer to where the data goes, and here it follows from the product's architecture rather than from a policy commitment. A hospital adopting this is largely extending software it already hosts.

Held at B rather than A for two reasons: no architecture description, hosting region or residency statement was located, so this is inferred rather than stated; and two newer components complicate the picture, since the mobile application moves data onto devices and the partner platform introduced in 2025 sits outside the record environment. A buyer should ask which components run inside the perimeter and which do not, because the answer has changed recently.

No pricing, pricing mechanism or contracting model located. Total funding located is roughly 7 million dollars across two rounds with the most recent in May 2022, which is small and now stale for a company that has just won its category's leading analyst award, and suggests either revenue financing or a capital structure not captured publicly.

One historical detail belongs on the record as a supplier durability observation rather than a criticism: the company launched its platform on a third party health technology marketplace in early 2022, and that marketplace's operator subsequently wound down.

Distribution channels in this segment have proved fragile, and a hospital embedding a vendor's software inside its record system takes on a dependency that is expensive to unwind, so the Behold.ai continuity question applies with force.

Specialty breadth is the explicit product thesis rather than an incidental property. The premise is that one record system cannot suit every acute care team, so the platform ships prebuilt workflows for many specialties and care team roles, with disease specific modules layered on including one for diabetes. Setting is inpatient acute care across the hospital, spanning physicians, nursing and multidisciplinary teams, with the mobile companion extending it away from a workstation.

Geography covers the United States, including a federal facility, and at least one large international academic hospital. Held at B rather than A because the coverage is deep within acute inpatient care and essentially absent outside it, with no ambulatory, post acute or community presence located.