Automated Claims Record Capture

Challenge  

• Inefficient, costly processing of incoming claims documents (e.g., summarizing, extracting records, classifying, etc.)
• Reliance on varying B2B solutions for receiving claims added complexity
• A scalable and accurate automation solution was needed to handle millions of documents annually

Solution  

• Implemented an automated system using AIP and GenAI to process 20M claims documents (200M+ pages) in 3 months
• Main optimization factors being cost and model accuracy during project scoping
• Achieved an exceptional accuracy rate of over 97% by surpassing the human benchmark

Results  

• Achieved a total processing cost of $300,000, equating to under $0.02 per document or $0.10 per claim on average.
• Unlocked significant value by enabling seamless use of extracted data and alleviating manual workload on the Claims Operations team
• Accelerated processing times, reduced leakage, and enhanced claims handling capacity