Bayesian statistical methods for insurance problems where data is sparse, prior knowledge is material or uncertainty quantification matters. From credibility models for emerging lines to Bayesian networks for claims causation — we apply the methodology where the problem demands it.