Oracle Spatial AI is a Python library deployed with Oracle Machine Learning (OML) and runs in Oracle Database spawned OML4Py servers in the Autonomous Database Serverless environment.

Spatial AI is integrated with OML, mainly OML Notebooks and OML4Py. It is therefore considered as an extension to the OML product with added spatial modeling capabilities.

The following diagram shows the architecture of Oracle Spatial AI integrated with OML on Autonomous Database Serverless.

Description of the illustration spatial_ai_architecture.png

You can interact with the database using the secure OML connection. Besides Autonomous Database, Oracle Spatial AI can also access spatial data from OCI Object Storage or the local file system.

You can access Oracle Spatial AI through the OML User Interface (see Get Started with Oracle Machine Learning User Interface). Using this interface you can perform the following:

• Create and manage projects and notebooks.
• Schedule and monitor model training and prediction jobs.
• Publish Python functions and models through the OML4Py Embedded Execution REST and SQL APIs to leverage the scalability and high-performance of Autonomous Databases (see Oracle Machine Learning for Python User's Guide).

See Get Started with Oracle Spatial AI for more information.