You did not lose the money.
Nobody looked for it.
Every billing company is paid a percentage of what it collects, which means it is paid to process claims and never paid to find the ones it failed to collect. Drop in your own A/R export and see the difference in ten seconds.
Drop your A/R aging export here
CSV from any practice-management system, or a raw 835 remittance file. Drop several files to analyze a whole group at once.
No signup. No account. Nothing is uploaded.
Export your A/R
Any aging or open-balance report. One row per service line, with the payer, the service date, the balance, and whatever follow-up notes exist.
Drop it above
It is parsed here, inside your browser. The file never touches a network and never reaches us.
Read the number
What is recoverable, which payer is quietly your largest, and what stops being collectible in the next 60 days.
Common questions
Is this really private?
Yes, and it is structural rather than a promise. The parser is JavaScript that runs in your browser tab. Your file is read from your own disk, analyzed in memory, and displayed. There is no upload, no server call, and no analytics event carrying your data. You can confirm it: open your browser network tab before dropping the file and you will see no request. Because the file never reaches us, running the diagnostic needs no business associate agreement. There is no way for this page to send it — the only thing that leaves is an email you write yourself, carrying totals and no patient detail.
What file do I need?
An A/R aging or open-balance report exported to CSV from whatever system you bill in. One row per service line. It needs a payer column and a balance column at minimum; service date and follow-up notes make the analysis considerably sharper. Raw 835 remittance files work too. If your headers are unusual, the parser will tell you which columns it could not read rather than guessing.
Where do the recovery estimates come from?
From a live A/R engagement, applied per tier rather than as one blended rate. We say exactly what the basis is on the report itself, and we tell you what your export did not contain. We do not publish benchmarks we do not have. As the corpus grows the ranges will be rebuilt on it, and we will say so when that happens.
What if the number looks too good?
Read the tiers rather than the headline. Recovery on aged A/R is genuinely partial, and the ranges reflect that. The estimate is most reliable where your export carried service dates and follow-up notes, and least reliable where it did not, which is called out in the report. The honest version of this diagnostic is more useful to you than a flattering one.
What does it cost?
We work the A/R on contingency: 30% of what is actually recovered, nothing on what is not, and nothing up front. No card is required to start and there is no system migration or change to how you bill today. Practices that want us to take over billing entirely move to 5% of collections after the recovery work proves out.