In insurance claims processing, unstructured documents – loss adjuster reports, emails, images- contain critical information buried across dozens of files. Extracting this data manually is slow and error-prone. In this video, Thorogood Consultant James Leonard and Trevor Hebron, Head of Data at Apollo Underwriting, explore how generative AI can transform this challenge.

Apollo built a modern data platform to solve their claims processing problem, but realized the real opportunity lay in their unstructured data. Watch how they developed a bespoke solution that balances innovation with control, delivering significant operational gains while maintaining the ability to iterate and adapt.

What We Cover:

• How to structure a modern data platform that supports both analytics and AI use cases
• Building a metadata-driven approach to document classification and field extraction
• Balancing off-the-shelf vs. bespoke solutions for generative AI in insurance
• Achieving accuracy and maintainability through business rules and language model tuning
• Scaling from pilot to production: processing thousands of documents through automated workflows

Is this For You?

• Are you responsible for claims processing or underwriting analytics?
• Are you interested in how generative AI could extract value from unstructured documents?
• Do you want to understand the data foundations required before implementing AI?
• Are you evaluating bespoke vs. off-the-shelf solutions for your organization?
• Are you looking to improve operational efficiency while maintaining control and agility?