Commercial banks receive numerous applications for credit cards. Several applications are rejected for reasons such as high loan balances, low-income levels or too many inquiries on an individual’s credit report. Manual analysis of these applications is mundane, error-prone and time consuming. Hence, this task of analysis and approval can be automated with machine learning (ML) algorithms. In this study, an automatic credit card approval predictor is built using ML techniques. The credit card approval dataset from the UCI ML Repository is used to train the ML model to predict if an applicant can be issued with a credit card or not. The performance of the ML model built using logistic regression is evaluated with and without grid search technique. It was observed that implementing grid search improved the competency of the ML model by 4%. Subsequently, an ANN model was developed. The ANN substantially outperformed both the logistic regression models. This study indicates that ML models can be utilized to solve complex problems which are arduous to decipher with the aid of conventional programming techniques.