AI leverage for every location — with patient data handling designed for teams accountable to payers and patients.

Cartulary gives dental and medical groups one control layer for AI-assisted administrative work: location-level access, human approval, and an audit trail for each patient record touched.

01

Workflows under practice control

ENTRY DSO-001
Review gate

Claims & RCM follow-up

Draft payer follow-up notes, missing-document requests, and denial summaries without letting staff paste PHI into unmanaged tools.

Human approved

Recall and reactivation outreach drafting

Generate location-specific recall and reactivation drafts from approved templates, then route them to the front desk for approval.

Masked context

Referral & treatment-plan document prep

Prepare referral packets, treatment-plan explanations, and handoff notes with record access scoped to the right practice role.

02

One recall workflow, written down

ENTRY DSO-014

A location manager starts with an inactive-patient segment. Cartulary applies the group's approved outreach policy, drafts the message, masks restricted details where the policy requires it, and queues the draft for front-desk review.

  1. Request opened. who asked · location · patient segment · policy version
  2. Data scoped. front-desk role · recall fields · restricted
  3. Draft prepared. template · model route · review queue
  4. Human approval recorded. approver · timestamp · patient record touched
03

Control layer proof

ENTRY DSO-027

Permissions follow the group structure: region, location, role, and record type. Every action is tied to an audit entry a compliance, RCM, or operations lead can review.

04

Compliance posture

ENTRY DSO-041

Designed to support HIPAA-regulated workflows; BAAs where applicable; PHI never used to train vendor models where contractually supported.

location-level permissions · audit trail per patient record touched · human review before patient-facing use

05

Founder-led implementation

ENTRY DSO-052

The first working session is with the person building Cartulary. Bring one workflow, one location policy problem, and one patient-data constraint. The output is a practical control map for your group.

Working session

Map one administrative workflow before it spreads across every location.

Use the session to decide what gets automated, what needs human review, and what patient data cannot leave your policy boundary.

06

Practice AI Policy Template (front-desk edition)

ENTRY DSO-063

Give front-desk teams a plain policy for patient data, approval steps, and AI-assisted drafting before tools appear location by location.

No patient data. Use a work email so the template can be tied to the right vertical.

07

FAQ

ENTRY DSO-074
Will practice-level teams actually adopt this?

The first workflows are already familiar: recall drafts, payer follow-up, referral packets, and treatment-plan prep. Cartulary adds policy, approval, and logging around work teams already do.

Does this require a deep PMS integration before we see value?

No. The initial map can start with document and queue-based workflows. PMS integration can follow once the group has decided which data fields, locations, and roles should be in scope.

How do you handle PHI and vendor model training?

The control layer is designed to route work under written policies, BAAs where applicable, and vendor terms that keep PHI from being used to train vendor models where contractually supported.

Who reviews patient-facing drafts?

Your group decides. Most teams start with front-desk leads, practice managers, or RCM supervisors approving messages before patient or payer use. The approval is attached to the audit trail.