News | June 5, 2023

Roots Automation Introduces InsurGPT - The World's Most Advanced Generative AI Model For Insurance

Accelerates the adoption of AI across the Insurance market

New York, NY /PRNewswire/ - Roots Automation, creator of AI-powered Digital Coworkers, today announced the release of InsurGPT - the world's most advanced fine-tuned Generative AI model, designed specifically for the Insurance market.

InsurGPT further expands the natural language capabilities of Digital Coworkers by utilizing proprietary, fine-tuned Large Language Models (LLMs) to read and accurately extract data across:

  • Structured documents - ACORD forms, medical claim forms, etc.
  • Unstructured documents - insurance applications, loss runs, quotes, first notices of loss, demand packages, general correspondence, etc.

InsurGPT utilizes a deep corpus of insurance specific data and documents alongside systems and process knowledge to significantly improve the accuracy, speed, and validity of data extraction and inferencing, while reducing false-positives, 'hallucinations,' that non-industry specific Generative AI models produce.

InsurGPT is designed by Insurance operators for Insurance operators and is natively embedded in all of Roots Automation Digital Coworkers. Critically, each interaction between a human and Digital Coworker improves the InsurGPT model, benefiting all customers.

"We have fundamentally imbued our Digital Coworkers with deep insurance experience and knowledge through InsurGPT" comments Chaz Perera, Co-founder and CEO of Roots Automation. "InsurGPT removes the need for templates or heavy OCR configuration to analyze and process business documents, it simply reads insurance documents like a human would and produces accurate results, fast and at scale."

Importantly, Roots Automation Digital Coworkers utilizing InsurGPT are already delivering real results for Insurance operators.

For example, a national Third-Party Claims Administrator (TPA) is using InsurGPT to scale its First Notice of Loss (FNOL) process.

This TPA was challenged with standardizing claim setup and document handling processes while servicing the varying needs of carrier partners and customers. 

Traditional hard-coded rules, decision logic, and templates failed to effectively factor for document complexity and the idiosyncrasies of carrier partner/customer special account instructions.

Leveraging InsurGPT allowed the TPA to work with non-standard instructions and documents within the FNOL process, resulting in greater accuracy without requiring a change to the underlying process.

"Critical to the development of InsurGPT was the fundamental belief that while companies like OpenAI and Google - and the entire open-source community - have created something great, they are not interested in going deep into the Insurance sector - nor have they considered the security, data privacy or heavily regulated nature of the insurance space. We have!", comments John Cottongim, Co-founder and CTO of Roots Automation. "Importantly InsurGPT is constantly improving through the real-time interactions between our Digital Coworkers and our Customers' teams".

InsurGPT is the most advanced insurance-specific Generative AI model which enables organizations to remove routine tasks, increase digitization and accelerate the adoption of AI.

About Roots Automation
Roots Automation combines machine intelligence and human ingenuity to create intelligent Digital Coworkers, providing organizations with AI-powered, digitized employees that can think, read and intuit like people.

Digital Coworkers are pre-trained to understand and interact with documents, systems and processes commonly found in insurance.

Digital Coworkers are always on, ultra-secure and deliver ROI from day one - freeing a human workforce of inefficient, soul-destroying work, increasing their productivity and job satisfaction. Roots Automation is based in New York and was founded in 2018. For more information, visit https://www.rootsautomation.com/.

Source: Roots Automation Inc

Copyright 2023 PR Newswire. All Rights Reserved