FAQ

Is this automated coding?

No. The decision core is deterministic rules and people. The pipeline prepares bindings and codes, but anything disputed is blocked and goes to a human for authorization: the coder authorizes the choice between candidates, the billing office handles disputed act lines, the physician authorizes the note. The product makes human decisions fast and defensible; it does not replace them.

What does the insurer see?

For every contested line — the transformation chain in the minimum necessary scope: the source fragment, the term, the descriptor, the code, the tariff. Not the patient chart and not the full document. Both sides see the same thing — that is the point of the reconciliation panel.

Where does our data go?

Nowhere. Processing runs inside your organization’s perimeter; in the cloud option — in the protected HealthOS contour located in Russia. No trainable models sit in the decision loop, and client data is not used to train third-party models. The calculator on this site computes on your device and sends nothing.

Is this a medical device?

No. The product’s purpose is administrative-financial: document structuring, coding, settlements. It does not diagnose and does not choose treatment; details — on the regulatory page.

We have our own clinical system and accounting software

Connection is via REST API or a standard export; no system replacement required. Documentation systems can use the standard agent profile. Integrations pass certification on reference scenarios.

What is the transformation chain?

Seven links from the source to a claim line: document → fragment → term → descriptor → code → tariff → claim, each with a version, an author and a cryptographic digest. The evidence is available on request. The full walkthrough is on the How it works page.

How is this different from other tools?

By properties, not promises:

• A coding assistant built on a language model produces a code but not the grounds for it — the argument with the payer falls back to an exchange of opinions. Here every line carries a chain in which each link has its own version and its own author.
• A code reference book is rows without a history and without anyone accountable. Here the code stands on a governed definition: a descriptor with a version and an author.
• A transcription system ends at the text of the note. Here the result is carried through to the personified service record and the claim line in the clinic’s accounting system.
• Rules hard-wired into clinical-system code are rewritten by a programmer for every regulation and every payer. Here the rules live in the governed graph and are versioned together with the vocabulary.
• Custom integrations are replaced by standard exchange: registry-bound structured documents and HL7 FHIR instead of an adapter per customer.

What does it cost?

Billing depends on volumes: per document, per physician workplace, a billing-office subscription, per verified claim for the insurer. We quote prices after a short interview about your volumes; pilot-program participants get special terms. Service details.

Why should we trust you?

You should not have to. Every result line opens, in one click, into its full chain; the version and author of every link are recorded; the audit log is immutable. We sell not a promise of quality but a way to check it.