A US insurance company was manually extracting provider names and charges from hundreds of inconsistent medical bill PDFs. Codoro built a desktop AI tool that processes all 800 client folders, handles documents up to 483 pages, and automatically produces client-level and company-wide Excel reports.
An unattended document-processing workflow replaced the manual review of medical bills across 800 insurance client folders.
Every billing cycle, employees opened medical bill PDFs one by one, searched for the provider name and total charges, and manually entered those figures into spreadsheets. The process had to be repeated across 800 client folders before the team could assemble a master company-wide summary. The difficulty was not simply the volume. Each clinic, hospital, and healthcare network used a different document structure, so rigid extraction templates could not process the full dataset reliably.
Provider names and total charges appeared in different locations, tables, sections, and page layouts depending on the healthcare provider.
Some files contained hundreds of pages, including a 483-page hospital record that could not be processed as a single AI request.
The application had to navigate the company's existing Client Files structure without requiring employees to reorganise or upload documents manually.
The company needed a provider-level breakdown inside every client folder as well as one consolidated summary covering the entire client base.
The workflow adapts to varying document layouts and automatically splits oversized files before extraction.
Codoro first mapped the company's SharePoint folder structure, the medical-bill naming patterns, and the exact fields employees were extracting. The automation was designed around the team's existing process so deployment would not require a new document-management system or a change in staff behaviour.
Instead of creating fragile templates for individual providers, Codoro connected the application to Claude through the Anthropic API. Each document is analysed for the provider name and every total-charges figure regardless of where those values appear on the page.
Documents exceeding 90 pages are divided into smaller chunks and processed sequentially. The extracted results are then merged into a single record, allowing the same workflow to handle both a 25-page clinic invoice and a 483-page hospital document.
The application generates a formatted provider-by-provider Excel breakdown inside each client folder and a consolidated master report on the user's Desktop. Once processing begins, no copying, spreadsheet preparation, or manual file placement is required.
800 client folders can now be processed in a single unattended run.
Manual opening, reading, and transcription of medical bill PDFs was eliminated from the workflow.
Documents ranging from 25 to 483 pages are handled without provider-specific templates or configuration changes.
Each client automatically receives a structured provider-by-provider Excel breakdown.
A consolidated master report is generated automatically for company-wide use.
The workflow scales with additional clients and documents without creating an equivalent increase in administrative work.
Staff can focus on insurance and legal work that requires human judgement instead of repetitive data entry.
Codoro AI delivered exactly what we needed, fast, precise, and without any hand-holding. Our team no longer touches medical bill PDFs for data entry. That work is simply done.
We map exactly where your operation loses hours and which constraint to enforce first — specific to your fleet, not a case study.