How was the process of deploying the model to production? Did you prepare documents and artifacts for MLOps, have access to production stands, and could you run pipelines independently?