×
From 60,000 manual claim entries
to autonomous extraction
Defined Physical Therapy manually extracted data from 60,000+ CMS-1500 insurance claim forms per year received via fax and email. A Nanonets agent now ingests from Gmail and Google Drive, extracts all claim fields via AI OCR, post-processes, and exports to their CRM end-to-end.
Defined Physical Therapy Network is a physical therapy practice based in Kennesaw, Georgia. Founded in 2017, the clinic processes tens of thousands of CMS-1500 insurance claim forms each year across its patient population — making accurate, timely claim submission critical to its revenue cycle.
The Problem
Defined Physical Therapy receives over 60,000 pages of CMS-1500 (HCFA-1500) insurance claim forms every year via fax and email. Each form required staff to manually download the PDF, read each field, and re-enter the data into the clinic's internal CRM — a repetitive, error-prone process with no automation.
- —11 minutes per claim — entirely manual field-by-field data entry.
- —~2 FTEs dedicated to claim processing, consuming significant payroll.
- —Error-prone manual entry led to rejected claims and billing delays.
The automation has transformed how we process insurance claims. What used to take our staff hours now happens in minutes.
The solution
One claim agent. From fax to CRM in under two minutes.
Nanonets replaced the manual loop with one autonomous agent. The agent orchestrates four sub-agents which each own one step of the pipeline — monitoring Gmail and Drive, classifying the claim form, extracting all CMS-1500 fields, post-processing and exporting — with no human in the loop.
Pipeline overview
Impact
AGENT ANATOMY
Nanonets AgentNanonetsNanonetsNanonetsSee it run on your process, with your documents.
Start free. No credit card. Or talk to our team about your workflow.