The problem
Anodyne makes diabetic therapeutic shoes. Medicare pays for a pair only when every form is complete and signed. That paperwork took 4 to 6 weeks, and most of the wait was for a doctor's signature.
A clinic starts a case on one of 4 intake paths. ACE checks it against 11 Medicare rules. The physician packet goes out by fax, with reminders until the signed forms come back. ACE reads the returned forms, and the case is ready to dispense. On the order side, Claude reads each purchase order, applies the customer's rules and puts it in a review queue. Staff post it to NetSuite.
ACE flags a missing foot measurement. The case goes back to the clinic before any fax goes out.
What we built
- ACE, an engine that checks each case against 11 Medicare rules before a form goes to a doctor. It names the exact fix, such as a missing foot measurement.
- A fax loop for the physician packet. It tracks each form, flags each reminder when it is due and logs every reach-out.
- Order entry. Claude reads each purchase order and applies the customer's rules. Staff check the queue before an order posts to NetSuite.
- A HIPAA setup. Claude can run on AWS Bedrock or Google Vertex, which both offer a BAA. CloudTrail keeps six years of logs, and patient data stays out of the code. These screens show demo patients.
The result
Anodyne processes orders on the system in production. The paperwork side is in final testing before it goes to clinics.
“They anticipated needs we hadn't even voiced. We knew it was working once we were live processing orders, and now we have real process automation across several business functions.”
AnodyneBobby