At CNA Insurance, I delivered a $10M premium benefit and improved the BOP liability loss ratio by 10% through a payroll exposure true-up. I also performed annual profitability studies across Property and Liability, covering the full ratemaking cycle.
I built and validated a two-variable Generalized Linear Model for loss ratio relativities, then reconciled the results to Tableau dashboards used for underwriting segmentation. My pricing work also included forecasting, rate filing support, rating-plan testing, and studies that informed underwriting strategy.
I designed the Small Business Agent, a multi-agent generative AI system that writes and runs SQL against BigQuery and Oracle and produces verified analysis deliverables. It won first place in CNA's department-wide Actuarial Agent-a-thon; I also created a Reserving Agent now piloted in reserve reviews.
Alongside this work, I migrated the Small Business pricing codebase to Google Cloud Platform and GitHub, converting more than 10,000 lines of R to Python and SQL. Earlier at CNA, I performed Specialty Reserving reviews and developed a probability-of-payment method for a lawyers professional liability program.

