The image shows a functional representation of the architecture. Oracle Cloud Infrastructure Data Integration provides data extract orchestration and data transform for data sourced from Oracle Fusion Cloud Enterprise Performance Management and other sources such as SAP and spreadsheets.

Oracle Autonomous Data Warehouse stores historical, incremental, and prediction data.

Oracle Cloud Infrastructure Data Science provides a machine learning feature engineering pipeline, a machine learning pipeline, and a scoring pipeline.

Historical data from Autonomous Data Warehouse is used in the machine learning feature engineering pipeline which provides data extraction, data preparation, feature engineering, and feature validation.

Data form the machine learning feature engineering pipeline is used by the machine learning pipeline which provides feature ingestion, model training, model evaluations, and model selection. Data scientists and data analysts using this data.

The scoring pipeline provides a model catalog API endpoint, model deployment functionality, and a machine learning model API endpoint. The diagram shows a functional path from Model selection in the machine learning pipeline to the model catalog API endpoint to model deployment, to the machine learning model API endpoint. The scoring pipeline scores new data from the Autonomous Data Warehouse incremental data, persists predictions data in Autonomous Data Warehouse, and feeds predictions sent to business users by using a chatbot. The chatbot also interfaces with Oracle Cloud Infrastructure Functions chat model and Llama 3.1 405b.