A lightweight, explainable “Epi-Copilot” that combines generative AI, biostatistics, and epidemiology to turn routinely collected, formatted primary care data (and related signals such as immunisation records, diagnoses, pathology trends, and medication changes) into actionable, interpretable risk insights for clinicians and epidemiologists. The goal is not a black-box predictor, but an explainable epidemiology system that supports proactive care and safety surveillance.