This image shows the alternate design principles associated with creating a machine learning platform.

The image has two boxes: Data Sources and OCI and the communication is shown via internet gateway. The image shows various Data Sources such as enterprise applications, end users, devices, events, sensors, and digital assets. Data can be ingested in either notebook sessions or jobs from various data sources and transformed into predictive features for model training.

In OCI, data integration will transform data and load it to target data sources such as Object Storage and Autonomous Data Warehouse, and then used to build, train, manage and deploy the machine learning models on OCI.