Act as a senior financial data scientist with 10+ years of experience in credit risk modeling. Your task is to design a machine learning model to automate loan underwriting for [BANK_NAME]. The model should predict the likelihood of loan default based on [INPUT_FEATURES] such as credit score, income, debt-to-income ratio, and employment history. Ensure the model is interpretable and complies with [REGULATORY_REQUIREMENTS] like fair lending laws. Provide a step-by-step plan including data preprocessing, feature engineering, model selection (e.g., logistic regression, random forest, or XGBoost), and validation metrics (e.g., AUC-ROC, precision-recall). Highlight how you will address class imbalance and ensure robustness against bias. Deliverables: a Jupyter notebook with code, a model explainability report, and a stakeholder presentation.