M1 audit spike · Tested on synthetic data only
What does your plan’s payment history say?
Check vision-plan remittances against one practice’s own history. Review provisional underpayments, unresolved denials and patient balances with the observations behind every rate finding.
Evidence first. A ledger check second.
The audit learns fixed allowances, percentages and capped-charge rules by payer, employer group, code and modifiers. It checks the final third of the history against a frozen rule learned from the first two-thirds.
It does not declare money owed. Short or contradictory histories stay unexplained; a probable fee-schedule change is not an underpayment. Patient balances must be checked against the practice’s own ledger.
The sample is synthetic, not a customer result. Real-practice reproduction and uncollected dollars have not been validated. No quote screen, billing or hosted audit is available.
Nothing leaves the browser while scrubbing.
- Scrub all months together. Names and identifying segments are removed locally; claim numbers become unrelated random tokens.
- Keep the private key. Download the scrubbed remittances and a separate key file. Review locally; never send the key or original files.
- Review an offline report. The founder runs the audit only on reviewed, scrubbed files. Rules are never shared across practices.
This page has no upload endpoint, analytics, third-party assets or account. Use 835 files or the documented patient-free CSV; PDFs and scanned EOPs are unsupported.
Payment dates and group numbers are retained and have not had legal review. A structural gate is not a legal de-identification guarantee. The named partner must check the output before sharing; lawyer review is required before a second practice is invited.
A free audit for one practice
Design-partner offer: one named optical or optometry practice, at least six months of de-identified remittances for one non-VSP plan, and a joint check against its own ledger. No payment or account is needed.
Participation is arranged directly with the founder through your existing contact channel. There is no open intake form. Do not send any patient data before the local review is agreed.
The goal is to test whether history predicts paids and whether at least one finding is actually uncollected. If it does not, we stop rather than build a quote on unreliable history.
Start with the local scrubber