Bank credit risk models are important for assessing the likelihood of default by borrowers. Russian banks are moving towards using advanced internal rating models to determine regulatory capital. These models help banks plan strategically, simplify interactions with regulators, and potentially reduce capital requirements. The models predict default events based on credit policies and historical data. Different risk segments, like low-default assets and high-default corporate borrowers, are identified using economic and statistical criteria. The study describes how models are developed for various risk segments, including those with limited historical data like residential real estate and project finance.