Kind of, we started down the path of deploying with it, but realized that the number of moving pieces and chatter was too much. Ultimately decided to go for a managed solution. Databricks has a managed mlflow, sagemaker has an equivalent.