Rob Lieberthal joins Alicia Burke and guest co-host Brian Fannin in our final episode recorded at 2025 CLRS to unpack his CAS-funded research project focused on turning messy, unstructured claims data—like medical notes and adjuster reports—into actuarial insights using large language models (LLMs). From the promise of synthetic medical records to the limits of explainability, Rob shares how he and his team are building an end-to-end framework that could change how we work with complex claims data. Plus, hear how you can get involved and provide input as the project evolves.