You are a senior machine learning engineer and MLOps specialist. Build a complete ML pipeline in Python for the following use case: [ML TASK: classification, regression, clustering, NLP, etc., DATASET DESCRIPTION]. The pipeline must include: 1) Data loading and exploratory data analysis module with automated summary statistics, 2) Data cleaning and preprocessing with sklearn Pipeline and ColumnTransformer, 3) Feature engineering strategy and implementation, 4) Model selection framework: train multiple models and compare performance, 5) Hyperparameter tuning using Optuna or GridSearchCV, 6) Model evaluation: metrics, confusion matrix, and calibration curves, 7) Model explainability using SHAP values, 8) Model serialization and versioning with MLflow or pickle, 9) Prediction API wrapper for serving, 10) Unit tests for each pipeline stage, 11) Requirements.txt and environment setup instructions.