Recent advances in large language models (LLMs), such as GPT-4, have spurred interest in their potential applications across various fields, including actuarial work.
This paper introduces the use of LLMs in actuarial and insurance-related tasks, both as direct contributors to actuarial modelling and as workflow assistants. It provides an overview of LLM concepts and their potential applications in actuarial science and insurance.
In this webinar, we will:
Actuary GPT: applications of large language models to insurance and actuarial work (PDF. 680 KB) by Caesar Balona, Highly Commended paper, Brian Hey Award 2023
You can view the recording of this webinar on the IFoA’s Virtual Learning Environment (VLE). Watch recording.
Chair
Ritika Rustagi is an experienced general insurance risk and capital actuary, currently a Director at Mazars. With about 15 years in the industry, she has held various roles within insurance companies and consultancy firms working on capital models and validating them. She is excited about the opportunities AI brings for the insurance market.
Caesar is a qualified actuary with experience in the short-term consulting industry where until recently he headed up the data science team at QED Actuaries and Consultants. Currently, he leads Catastrophe and Climate Modelling at Old Mutual Insure where he uses machine learning and programming to gain insights on climate risk from remote sensing data.
A significant part of Caesar’s work lies at the intersection of actuarial science, software development, and machine learning. He’s created websites such as RiskyRouter, a route mapping software providing estimates of accident risks on travel routes within Seattle, which won the 2022 Casualty Actuarial Society Hacktuary Challenge.
More recently, he also developed Insurance Automata, an interactive application that models the spread of fire across a house, demonstrating the importance of adequate fire defences. Outside of these projects, Caesar contributes to the broader community through his website, modernactuary.co.za, and through research.