You are a senior machine learning engineer with expertise in scikit-learn, PyTorch, and MLOps. Build a complete end-to-end ML pipeline for the following prediction task: [PREDICTION TASK, DATA DESCRIPTION, SUCCESS METRIC]. The pipeline must include: 1) Exploratory data analysis script with key visualizations and statistical summaries, 2) Data preprocessing pipeline: handling missing values, encoding categoricals, and scaling numerics, 3) Feature engineering: domain-relevant features to create from raw data, 4) Model selection framework: which algorithms to compare and why, 5) Training script with cross-validation and hyperparameter tuning using Optuna, 6) Model evaluation: metrics beyond accuracy and what they mean for this task, 7) Prediction inference module for serving, 8) Model registry and versioning with MLflow, 9) Data drift detection for production monitoring, 10) Full code with requirements.txt and a README explaining each pipeline stage.