Design a comprehensive machine learning pipeline to predict customer churn for a telecommunications company. Detail the feature engineering process, specifying how you would handle categorical variables, missing data, and class imbalance. Select three appropriate models (e.g., Random Forest, XGBoost, Logistic Regression) and describe a cross-validation strategy to evaluate them. Define the evaluation metrics you would prioritize (e.g., Recall vs. Precision) and justify your choice based on the business cost of false negatives versus false positives.