This paper employs clustering and machine learning techniques to analyze validation reports. It provides insights into issues related to credit risk model development, implementation and maintenance.