Skip to content
← PlatformWorkflows & the sign-off gate

No clinical content reaches a patient until your physician signs it off.

A workflow is the clinical content the AI follows for a condition. We draft it from published guidance to save your team the blank page — but a supervising physician on your side reviews it, edits it, and signs it off before it ever runs. Until then, the system won't let it touch a real patient.

  • HIPAA-compliant
  • BAA-ready
  • CPT 99490-native
  • Transparent AI
Workflows · approval

Condition workflows

Each one needs a physician's sign-off before it can run.

Heart failure check-in

Draft built from published guidance. Covers between-visit symptom check.

Awaiting clinician sign-off
Can't be assigned to a real patient yet
COPD check-in

Reviewed and edited by your supervising physician, then signed off.

Approved & logged
Signed off · v3 · logged with name & date
The block is enforced in the database — not a setting someone can forget.
Illustrative · sample data
Read this first

No workflow runs with real patients until a supervising physician signs it off — enforced in the database.

To be plain about where we are today: the workflows in the product are drafts built from published guidance. They are scaffolding for a physician to review — not content any physician has approved yet. We won't claim otherwise, and the system won't run an unsigned workflow on a real patient.

What a workflow is

The clinical content behind a condition.

One workflow per condition. It guides what a check-in call covers and how the AI sorts what it heard. We describe its purpose here — the exact questions and the lines it draws stay inside the product, where they belong.

What the call should cover

For a given condition, the workflow guides what the check-in pays attention to — the topics worth asking a patient about between visits. The purpose, not the script.

How to sort what it heard

It defines how the AI triages a call into three buckets: urgent — page a clinician now; worth a look — into the review queue; routine — logged and trended.

Where the line sits

It sets the bar for what counts as a concern for that condition. Your supervising physician decides where that line is — and can move it.

A workflow is content, not code. That's the point: it can be read, edited, and signed off by a clinician without an engineer in the room — which is exactly what the sign-off gate is built around.

Physician sign-off gate

A draft becomes a workflow only when a physician says so.

The gate is the capability. It turns clinical content from something we wrote into something your physician owns — with a hard stop in between.

01

Drafted from published guidance

Each workflow starts as a draft built from published clinical guidelines — a structured starting point for review, never a finished standard of care.

02

Reviewed by your physician

A supervising physician on your side reads the draft, edits it to your standard, and signs it off. The clinician layer is a seam your medical staff fills.

03

Enforced, not promised

Until that sign-off exists, the workflow cannot be assigned to a real patient. The block lives in the database — it is a state the system enforces, not a policy we ask people to remember.

The supervising physician is on your side, not ours. You decide what the AI covers and where the line sits for each condition. It extends the reach of your care team — it never replaces a clinician, diagnoses, or prescribes.

A state the system enforces

The gate is a real status, not a promise.

Inside the product, every workflow carries an approval state. An unsigned draft sits in “awaiting clinician sign-off” and simply can't be assigned to a real patient. Sign-off flips it to “approved & logged” — recorded with the physician's name and the date.

Heart failure · history

Change history

Every edit and sign-off, with who and when.

Approved & logged
  1. Signed off — now live

    Dr. A. Rivera, supervising physician

    Jun 14 · 9:42 AM

  2. Edited the concern threshold

    Dr. A. Rivera

    Jun 14 · 9:31 AM

  3. Draft created from published guidance

    Family Guardian AI

    Jun 11 · 2:18 PM

History is kept, not overwritten — the version behind any call can be reconstructed.
Illustrative · sample data
Auditable by design

Who changed what, and when — kept on the record.

Because a workflow is content under your physician's name, the trail behind it has to hold up. Every edit and every sign-off is logged, and any call can be traced back to the exact version that ran it.

Every change is logged

Each edit to a workflow records who made it and when. The history is kept, not overwritten — so the version behind any call can be reconstructed.

Sign-off is on the record

Which physician signed which version, and on what date, is stored alongside the workflow. There's a name and a timestamp behind every approval.

Traceable back to the call

Any call can be traced to the exact signed version of the workflow that ran it — so a reviewer can always see the clinical content a patient actually heard.

The same trail underpins how we handle data and access across the product. If audit and control are what you need to see before a pilot, that's the right place to start.

See security & trust

Have your physician read a workflow. In the demo.

In a twenty-minute walkthrough on your panel, we'll open a real workflow, show the sign-off gate as the hard stop it is, and walk the audit trail end to end — the real product, not slides.

A real person replies — usually the same business day. No sales sequence, no obligation.