Meet the insurtech: DigitalOwl

Meet the insurtech: DigitalOwl

One night during a family dinner Yuval Man, a former lawyer, was telling his brother Amit about working on a medical review for personal injury claims and Amit said, “Machines can do it.” When Yuval went back to the office, he looked at some medical records and agreed that the repetitive tasks could be done by a computer. 

Yuval, co-founder and CEO of DigitalOwl, called Amit, co-founder and CTO, and told him they should start a company. They focused first on legal automation, later realizing that the insurance industry had a wider use case for reviewing medical records. 

Yuval Man

DigitalOwl has partnerships with Reinsurance Group of America and Nationwide. 

Yuval said the technology mines medical information. “It extracts medical data from the source document to generate the full medical story of a patient in a paragraph, or give a summary of a specific document.”

DigitalOwl has two proprietary AI technologies. An entity extraction engine, which is a natural language processing (NLP) engine that extracts information from source documents. The extracted data points are enhanced by the company’s proprietary medical knowledge base, adding medical codes, medical hierarchy, severity, co-morbidity relations, etc.

The company also has a large language model (LLM) that was trained and fine-tuned specifically for medical/insurance use cases. The engine reads source documents and generates new text that gives a full medical picture. It includes a paragraph that summarizes the entire case, a specific document, or a specific impairment.

Yuval said the company can solve hallucinations using the entity extraction engine in combination with Gen AI to flag information. They also have a large team of medical experts, underwriters and claims adjusters who train the engine on medical information.

See also  Can gen AI teach the next generation of underwriters?

DigitalOwl solves the black box issue by making all the information in solutions clickable. Click on it and see the source document where it came from.  “Everything is clickable. … Click on the information and see the source document,” Yuval said. 

Yuval said he doesn’t see the technology replacing humans but acting as a copilot. 

“The main product we have is something that was meant for humans to review,” he said. “An underwriter or a claims adjuster will get this summary, … and be able to dive in and search.”

The summarization and compiling of information removes about 80% of the repetitive work, said Yuval, and it frees up underwriters and adjusters to do the art of their job, he added. 

Yuval said technological advancements are accelerating with Gen AI. “It’s like the invention of the internet. … Soon we will be in a place where you can access an AI tool that is like having the smartest underwriter in the world next to you and you will be able to talk to them like a human.”