Natural language learning speeds up medical claims processing

Industry : Cross-Industry, Insurance

Function : Operations

Company Profile

The Hartford Financial Services Group, Inc., usually known as The Hartford, is a United States-based investment and insurance company. The Hartford is a Fortune 500 company headquartered in its namesake city of Hartford, Connecticut. It was ranked 160th in Fortune 500 in the year of 2020. And had $20.74 Billion in Revenue in 2019. The Hartford is the 13th-largest property and casualty insurance company in the United States. It sells products primarily through a network of agents and brokers, and has also been the auto and home insurance writer for AARP members for more than 25 years.

Scope Highlights

  • Natural language models and learning
  • Tableau
  • Integration with multiple systems and data stores
  • Medical claims
  • Intelligent process automation


The client, a nationwide leader in financial services, processes a high volume of complex medical claims. This process involved 100 nurse practitioners manually reviewing over 5 million medical records per year to confirm or update details in the claims. Some of these medical records were up to 50 pages long and took 40 minutes per record to review and make comments. Leadership wanted a scalable solution that would let team members focus on more complex claims details.


NLP capabilities automated processing of medical record images across multiple lines of business. The solution uses trained natural language models to extract, classify and standardize key elements from each record such as ICD codes, body parts, treatments and more. The solution provides a web application experience integrated with the existing claims systems to allow authorized individuals to review details, provide model training feedback and decision claims.


$2.4 million annual savings

Decreased claims processing time by 40%

Saved 16.5 FTEs per year