This architecture demonstrates the abilities of a cloud data lake using Oracle Data and AI services to process data from a broad range of enterprise data resources. Use this architecture to leverage the data for business analysis and machine learning. Data can be discovered from various sources both within and outside Oracle and then ingested for transformation. Transforming the raw data to measurable and actionable data requires the data to be processed through various stages:

1. Discover
2. Ingest
3. Transform
4. Curate
5. Analyze, learn, & predict
6. Measure & act

The following features span across the entire Oracle Cloud Infrastructure (OCI) region:

• Governance
• Security, Identity, & Access Management
• On-Premises FastConnect enables online data transfer with Flexible Compute and Flexible Storage features of OCI
• Oracle's partnership with Azure Interconnect enables migration to the cloud and building cloud native applications.

The region is divided into functional layers that house physical or functional components:

• Discovery stage: You can discover data from a wide range of sources such as Enterprise Applications, Azure and other databases, Oracle Data Cloud, Webclicks, Event Streams, Sensors, and Media, or File Object Stores.
• Data Refinery layer: The data goes through the Ingest stage in this layer.

  • Batch Ingestion (OCI Data Integration, ODI, and DB Tools) services consumes application data. Oracle Autonomous Data Warehouse consumes refined application data. The Cloud storage in the Data Persistence Platform layer consumes raw data.
  • Change Data Capture (GoldenGate and ODI) and Bulk Transfer (FastConnect Data Transfer, MFT, CLI) enable raw data to move from all data refinery paths to cloud storage in the Data Persistence & Processing layer.
  • Streaming Ingest (Streaming Service, Big Data Service) service consume the Event Stream data. The Streaming Processing (GoldenGate Stream Analytics) service consumes the stream data and transfers it to the cloud storage in the Data Persistence & Processing layer.

• Data Persistence & Processing layer: Data goes through the Transform and Curates stages. This layer facilitates data navigation to show the current business view.

  The Data Persistence & Processing layer structures the data based on whether the database technology you use is relational or non-relational. Governance (Data Catalog) applies to application data and raw data when they pass through this layer.

• Access & Interpretation layer: Data goes through the Analyze, Learn, and Predict stage. This layer makes the data ready for access and intepretation using Analytics, Machine Learning, and AI Services such as Anomaly Detection. You can visualize refined application data using the Oracle Analytics Cloud service. Data scientists can leverage the Machine Learning (Data Science, OML Notebooks, OML) services to build and train models with a familiar user interface. Machine Learning consumes raw data which can be used for training models. Streaming Analytics (GoldenGate Stream Analytics) services provide data visualizations to make the data available for access and interpretation.

  The APIs available through an API Gateway and Functions can be used by developers to build their own applications and leverage the raw data using machine learning and AI services.

• Measure & Act stage: Oracle Applications Data Warehouse can leverage the analyzed data, use it, learn from it, and predict outcomes. Augumented Analytics, Dashboards & Reports, Machine Learning Models, Data-Driven Applications, AI Enabled Services and all benefit by using the measurements and acting on the predictions. Organizations can monetize the data by making data-driven business decisions using data-driven apps. They can train machine learning models, build dashboards & reports, and augumented analytics.