Credit Risk Predictive Modeling Automation Pipeline

Design an end-to-end machine learning automation pipeline for continuous credit risk assessment. Implement a system that: 1) Continuously ingests alternative and traditional credit data sources, 2) Applies ensemble machine learning models for risk prediction, 3) Generates dynamic credit scoring models, 4) Provides real-time decision support for loan underwriting, 5) Includes model drift detection and automatic retraining mechanisms.