Loans make up significant part in bank profits. In a bank, amongst large number of loan applications, it can be lengthy and challenging to choose genuine and eligible applications to approve the loan, if the process is done manually. The system was developed using Python 3.7 & its libraries and Jupyter Notebook cross-platform Integrated Development Environment (IDE). The developed system uses the machine learning algorithms - Naïve Bayes, Support Vector Machine (SVM) and XGBoost (Extreme Gradient Boosting) individually to classify the loan applications to automatically choose genuine and eligible applications. Based on average bank balance of the period and financial background, the system approves the loan amount to the predicted applicants with good credit history. The system’s performance in predicting the credit worthiness of applicants was evaluated. Naive Bayes algorithm predicted the credit worthiness with 79% accuracy, SVM algorithm with 81% and XGBoost algorithm with 84%.