How I count and verify clinic findings.

I find revenue that clinics are missing, by reading their own Jane.app data. Numbers like that are easy to claim and hard to check, so this page shows the counting.

Every public number I use is held to the standard on this page. If a number cannot meet it, it comes down until it can. That has already happened once.

The note where I took my own number down →
One signal's journey
  1. 1DetectedSignal detected
  2. 2EvidenceEvidence reviewed
  3. 3OwnerOwner assigned
  4. 4ActionAction completed
  5. 5VerifiedOutcome verified

What gets read.

Caretrics reads twelve months of a clinic's Jane.app data with read-only access and works between Booked and Paid. Signals come from the clinic's own calendar, visit, and billing records, never from clinical judgement. Access can be revoked anytime.

Read, not inferred.

Each item links back to the visit it came from, the date it was submitted, and the last status recorded. Nothing is inferred. Where a field was missing, the item says so instead of guessing.

A finding needs an owner.

The clinic assigns each finding to a person. Caretrics does not decide who. It records the name and the date, so the work stops being nobody's in particular.

Identified is not collected.

A scan identifies money that may be slipping. Nothing counts as recovered until it lands. The proof is what changed afterward: the claim that went out, the patient who rebooked, the authorization renewed before the visit.

How an outcome closes.

An item only closes when its status changes in Jane. Identified and collected are reported separately, and the record shows plainly which of the two happened.

Where AI assists, and where a person decides.

Anything automated says what it read, what it skipped, and what it is unsure about. If it cannot, it does not get to make the call. People assign the work and make the decisions, and each action is logged against the item, never against the person who made it.

Public examples.

Patient examples use a first name and last initial, never real data. Signals shown in public are anonymized, and every public number carries what it counts and where it came from. When a number is missing its breakdown, it comes down.

The $127,000.

The number on the homepage: one clinic, 25 providers, the twelve months of June 2025 through May 2026. Four gaps: unbilled claims, $52,000. No-shows, $37,000. Missed re-bookings, $24,000. Missing re-evaluations, $14,000. Identified across those four, and identified is not collected: it is what the records showed, not what the clinic banked.

  • Unbilled claims$52,000
  • No-shows$37,000
  • Missed re-bookings$24,000
  • Missing re-evaluations$14,000

One clinic, 25 providers, June 2025 through May 2026. Identified is not collected.