Whenever you're ready, walk me through how you'd approach this.
Describe how you would build an end-to-end pipeline that performs joins, filtering, null imputation, encoding, scaling, rounding, and then trains a model to predict price from transformed features.
Explain the pipeline stages, data contracts, train-serving consistency, orchestration, reproducibility, and validation strategy. Include how you would handle failed records, schema changes, reruns, feature leakage, and monitoring in a practical implementation.