These days I meet many people who talk about how they have a project to find an AI-plattform to use on-premises. Using it in the cloud on Autonomous it is pretty obvious that you use one in a handful of AI-services Oracle has at your disposal there. The question is what you should use to build AI-services on the data in your on-prem Oracle database.

You don’t need to get anything.

This blog post assumes you have an Oracle database 18c or later. If not it may serve as motivation to upgrade.

Oracle has Machine Learning libraries available for R, Python and more. But also in the database. There are many values with it but one key benefit is that you do not have to bring data to your AI platform. You do not have to bring algorithms to your data either. It is already there for you and it is included in your license.

What you do need is of course to learn how to use it. The easiest way to get started is to use the Data Miner graphical component you also have access to. You can built workflows, look at ML-results and more directly in SQL Developer with the Data Miner component.

To get started, here are som links:

Overview:
https://www.oracle.com/database/technologies/datawarehouse-bigdata/oml4sql.html

Documentation (select your version – from 19c has more complete docs):
https://docs.oracle.com/en/database/oracle/machine-learning/oml4sql/index.html

Short course top get started using Data Miner in SQL Developer:
https://apexapps.oracle.com/pls/apex/f?p=44785:50:14439591210609:::50:P50_COURSE_ID,P50_EVENT_ID:457,6310