A multi-finance institution, also known as a non-banking financial institution (NBFI), provides a range of financial services and products beyond traditional banking services. While banks primarily accept deposits and provide loans, multi-finance institutions focus on offering specialized financial services such as consumer financing, leasing, factoring, insurance, and other credit-related activities. PT XXX is a 31-year-old comprehensive multi-finance company with a stable credit rating. However, PT XXX is still struggling on identifying potential customers and assessing credit applications effectively. This study aims to solve the problem by finding the best model by comparing popular suitable model based on literature review with machine learning technique to minimize the risk of default. dataset contains 252.000 rows and 13 columns including risk_flag as target variable. The study started on obtaining the necessary data, data cleansing from null data, Exploratory Data Analysis (EDA) conducted with statistical analysis to understand the dataset, building the model equipped with SMOTE on the dataset, 30% data testing and 70% data training partition followed by metrics evaluation on accuracy and recall. Random Forest become best model in this study with 89.22% accuracy and 77.68% recall.