You have developed a new ML model in a Jupyter notebook and want to create a reliable, repeatable training process that tracks model artifact versions and lineage. You plan to retrain the model weekly. How should you operationalize this training process?