Schedule of CPD: AI Tools for Actuaries (14/01/2027) - 2 CPD

From statistics to AI 
A short overview of how the actuary’s statistical toolkit extends into machine learning, language models and generative AI. 

Predicting Initial Claim Reserves with Machine Learning 
When a claim is reported, the insurer must set an initial reserve, often using simple rule-based approaches. This use case uses motor casco (own-damage) claims data to show how XGBoost with a Gamma deviance loss can predict the final claim cost from the information available at notification. It also covers practical data issues and tests whether claim descriptions processed with language models add predictive value. 

Extracting Knowledge from Actuarial Code with Generative AI 
Actuarial models often sit in poorly documented legacy scripts. This use case shows how large language models can summarise model logic, extract hard-coded assumptions and generate documentation. 

Combining ML Models for Email Attachment Classification  
IF TIME ALLOWS 
ML models are often combined into powerful customised tools. This case study shows computer vision, OCR and text classification working together in practice.