Credit risk assessment of individuals is traditionally determined using data such as age, income, education level, and FICO or Equifax credit scores. The increasing popularity of e-commerce and the large amount of bank card transaction data that can be collected and associated with clients these days suggest that a machine learning approach that incorporates this data can be useful for credit risk determination. Using real world data about bank card transactions and default history of other clients, we built a model to assess the credit risk of an applicant that requires only their bank card transaction history.