Automated Student Performance Predictive Analytics Pipeline

Design a comprehensive Python data pipeline using pandas and scikit-learn that automatically ingests student academic records from multiple sources (CSV, SQL databases, API endpoints), performs multi-dimensional performance prediction, and generates customized intervention recommendations. The system should include automated feature engineering, handle missing data, implement cross-validation, and produce a machine learning model that predicts student risk of academic failure with at least 85% accuracy. Include robust error handling, logging mechanisms, and a Flask-based dashboard for administrative visualization.