How a Fortune 500 insurance company saved 30% manual labor by automating its quoting workflow

Challenges

• Each carrier uses their own jargon and format
• Many provisions need to be interpreted, not just extracted
• There can be multiple employee classes of coverage
• Plans frequently have complex tables with multiple entries per row
• Multiple products can be intermixed within one plan

Because of this difficult context, conventional tools failed to extract accurate information from insurance plans, forcing the company to rely on expensive manual labor.

Solution

Due to its sophisticated natural language processing, SemanticPro was able to solve the above mentioned challenges while producing highly accurate results. The solution was trained on five different products: Long-Term Disability (LTD), Short-Term Disability (STD), Vision, Dental and Life Insurance.

The solution gave full control to Subject Matter Experts (SMEs) over the whole extraction process, giving them the possibility to fine-tune the system without the intervention of any AI expert.

By using SemanticPro, the company was able to automatically:

• Extract key information from five types of insurance plans from other carriers

• Interpret this information into the company’s language

• Accurately extract key information from complex tables with multiple entries per row

• Detect employee classes and associate extractions to class description

• Classify clauses, even when the wording differs

• Detect and OCR scanned documents without manual intervention

• Export the document extractions in an Excel format

The Cortical.io Difference

• The solution extracts information and interpret provisions with a very high level of accuracy
• Extraction models are trained quickly, based on annotations by the firm’s subject matter experts on not more than 100 sample plans
• The subject matter experts have full control over improving the process by verifying, correcting, and adding extraction targets
• No need for AI experts nor data scientists