Automated Credit Scoring Machine Learning Pipeline

Create a comprehensive JavaScript machine learning pipeline for automated credit risk assessment using TensorFlow.js. The system should ingest multiple financial data sources, perform feature engineering on credit history, income, and behavioral metrics, and generate probabilistic default risk scores. Implement cross-validation techniques, support model retraining workflows, and develop an interpretable dashboard showing feature importance and model confidence intervals.