MLOps works with a AWS cross-account role that enables us to read your datasources, create AWS Sagemaker and AWS Glue jobs. You can inspect the full Cloudformation stack used on your account here.
Furthermore, the following applies to all MLOps users:
MLOps will not save any of your datasources or any other customer specific information unless an explicit permission has been given. Our goal will always be to save the minimum amount of data on you for having an operational platform.
2. MLOps will save metadata on all customers interaction with the platform such as preprocessing scripts, training scripts, datasource paths, metrics on model performance etc. Exactly what data we have stored on you can be accessed at any time by emailing firstname.lastname@example.org.
3. MLOps will never sell or provide any of your data, metadata or other information to a third-party provider without your explicit consent.
4. When you use the MLOps platform, you pay for the SaaS application and all it's provided accessories. The price you pay for this does not include the costs for training and deploying machine learning models on AWS Sagemaker or creating datasets on AWS Glue with Spark. There will also be an insignificant extra cost on your account based on the amount of usage due to the fact that we collect events via CloudWatch.