The challenge
When electronic prescribing fails, prescriptions still arrive by fax, and a $400M high-volume pharmacy was processing them by hand. Standard OCR couldn’t read the unstructured fax images, so technicians re-keyed each one, at about ten minutes per prescription.
That meant high labor costs, transcription errors, compliance risk and delays for patients.
Our approach
We deployed AI automation that uses multimodal models to read faxed prescriptions straight from the image and turn them into structured data, with no templates or setup, across thousands of fax formats. It extracts the key fields (patient, drug, dosage and prescriber), validates them and passes them into the pharmacy management system.
The outcome
Processing time fell from about ten minutes per prescription to seconds, saving about $500K a year in labor and shipping costs, with a return on investment in under three months. Transcription errors were eliminated, and technicians moved from data entry to patient care.
to process a faxed prescription, down from about 10 minutes
to return on investment
AI maturity, up from Level 3
