Case Study - Classifying millions of incoming claims across hundreds of categories
09 August 2019
DIGITAL MAILROOM AUTOMATION AT SCALE
INTRO
The client is a massive business process management company with its operations in U.S., U.K., India and Philippines.
They offer solutions for banking and financial services, customer services and healthcare operations to their clients including Fortune 500 companies.
PROBLEM
One of the largest American Health Insurance companies receives a large volume of claim applications in varying formats from a wide range of channels. They approached our client to build process automation for classifying their claim applications.
Classifying and extracting data from nearly 50,000 claim communications each day was slow and expensive. Manual handling of these claims was taking up a lot of people time and leading to errors that were costing millions in penalties.
PAIN POINTS
- Scanned images each comprised of 30+ claims were received through multiple interaction channels. These needed to be sorted across 200+ claim categories.
- The task of claim classification demanded in-depth reading and understanding of claims. Prior automation efforts lacking contextual awareness failed miserably.
- Manual sorting by 1000+ people was not scalable. More volume meant more people, leading to more errors and inefficiencies.
- Outsourcing highly sensitive individual health record data for automation was causing legal difficulties.
Make your mailroom process smarter and more scalable