The Oracle AI Data Platform helps organizations unlock the value of their data with a modern, AI-ready foundation. It unifies data storage, cataloging, ingestion, preparation, and analysis in one platform, enabling teams to turn raw data into insights faster.

Users can quickly create secure AI environments, build rich catalogs, manage access with RBAC, and explore data using Spark-powered notebooks. With Oracle Analytics Cloud integration, business users can instantly analyze curated datasets, providing a single, unified view of the data and AI lifecycle—backed by Oracle’s security and reliability.

In this blog, I will cover how to provision and get started with Oracle AI Data Platform . This article walks you through user access management, workspace and cluster setup, data catalog setup, notebook management, and concludes with a simple medallion architecture implementation. To make it easier to follow along, I’ve included detailed videos with voiceovers for each setup step. The topics covered in this article are:

AIDP Provisioning

User Access Management

Workspace and Cluster Setup

Setting Up Catalogs with OCI Object Storage Data

Setting Up Catalogs with Oracle Autonomous AI Database

Notebook Management: Export & Import

End-to-End Medallion Architecture Implementation Example

Oracle AI Data Platform Provisioning

To provision Oracle AI Data Platform:
1. Log in to the OCI Console.

2. Navigate to Analytics & AI and select AI Data Platform.

3. Click Create AI Data Platform.

Enter the following details:

Name: Provide a meaningful name for your Oracle AI Data Platform instance (for example, myaidp).

Description: Optionally, add a brief description to help identify the instance.

Workspace Name: Specify a name for the default workspace (for example, myworkspace).

Workspace Description (optional): Provide a short description of the workspace.

4. Next, under Access Level, choose either Standard or Advanced. Your selection determines how Oracle automatically generates the required IAM policy statements for your tenancy or selected compartment.

5. Click Copy to copy the generated policy statements and apply them in your OCI tenancy or compartment.

6. Once the required policies are configured, return to the provisioning screen.

7. Click Create to provision the Oracle AI Data Platform instance. After the provisioning process completes, the Oracle AI Data Platform instance appears in the console.

8. Click the instance name to open it and start using Oracle AI Data Platform.

For additional details and a step-by-step explanation of the provisioning process, refer to the accompanying video with voice guidance below.

🎥 Video: AI Data Platform Provisioning

User Access Management

Managing user access in Oracle AI Data Platform is simple and ensures that the right people have the permissions they need. To provide access:

From the Navigator menu, select Roles to view System and Custom Roles.
Click the desired role name,for example, AI Data Platform Admin Role.
Select the Members tab and click Add Member.
Set Principal Type to User and in Add by, select User Name or OCID.
Select the Parent Compartment and appropriate Domain.
Click All Identity Users, enter the user’s initials, and press Enter.
Select the user and click Create.

The user is now added with the assigned role and can access the platform.

For more details and a step-by-step walkthrough of granting user access, refer to the accompanying video with voice guidance below.

🎥 VIDEO: ▶ AI Data Platform User Access Management

Workspace and Cluster Setup

In Oracle AI Data Platform ,a workspace is an isolated environment used to organize notebooks, workflows, and libraries. You can create one or multiple workspaces based on project or team requirements. To create the workspace:

From the e Navigation menu, select Workspaces, and click Create.
Enter a Workspace name and Description.
Choose a Default Catalog (or keep the default). This is the default catalog used by the workspace to retrieve data.
Click Create to provision the workspace

A compute cluster provides scalable CPU/GPU resources for data processing and ML workloads.Clusters are required to run notebooks and Spark jobs. To create the cluster:

Navigate to Compute and click Create.
Enter a Cluster Name and Description.
Set Runtime Version to Spark.
Select Driver and Worker shapes.
Specify the Number of Workers or enable Autoscaling based on workload requirements.
Set the run duration or idle timeout.
Click Create.

Note: You can select Quickstart as a configuration option to rapidly provision a compute cluster. Quickstart offers a preconfigured setup designed for fast startup and everyday workloads, with sensible defaults for CPU cores, memory, and autoscaling. This allows you to start running workloads immediately without spending time on detailed cluster tuning.

Once your compute cluster is set up, attach it to your notebook before running any code:

• Open your workspace and create a new notebook.

• Attach your compute cluster to the notebook before running any code.

• Once attached, your Oracle AI Data Platform environment is ready to use for development.

For a detailed, step-by-step guide on creating workspaces, setting up compute clusters, and attaching them to notebooks, watch the accompanying video with voice instructions below.

🎥 VIDEO: ▶ AI Data Platform Workspace and Cluster Setup

Setting Up Catalogs with OCI Object Storage Data

In Oracle AI Data Platform, you can create standard catalogs or external catalogs. Standard catalogs are created and managed within Oracle AI Data Platform. External catalogs reference data stored outside the Oracle AI Data Platform, allowing you to access external data sources seamlessly. To create a standard catalog:

From the Navigator menu,select Master Catalog.
Click Create Catalog in the top-right corner to open the catalog creation page.
Enter a Catalog Name and Description.
Set Catalog Type to Standard and select the appropriate OCI compartment.
Click Create. After a few minutes, your catalog will be ready.

In data catalogs, a schema is a construct used to organize data. Catalogs serve as logical containers for schemas (also called databases), which can include tables for structured data and volumes for unstructured data. To create a schema:

• Open your newly created catalog, and click Create Schema.

• Enter a Schema Name and Description.

• Click Create. Your schema is now ready for organizing data volumes.

Volumes store data in its native form, supporting both semi-structured and unstructured data. To create a volume:

Navigate to your schema and select the Volume tab.
Click Create Volume in the top-right corner.
Enter a Volume Name and Description.
Set Volume Type to External.
Select the Compartment and the Bucket where your data files are stored.
Choose the specific Folder containing your files.
Click Create.

Your volume is now created and linked to the data in your OCI Object Storage bucket. You can open the volume to view the files stored in the selected folder.

For a step-by-step walkthrough of creating catalogs using OCI Object Storage data, watch the accompanying video with detailed voice instructions.

🎥 VIDEO: ▶ AI Data Platform Setting Up Catalog with OCI Object Storage

Setting Up Catalogs with Oracle Autonomous AI Database

External catalogs enable Oracle AI Data Platform to access data from sources like Oracle Autonomous AI Lakehouse, Oracle Database, or Autonomous Transaction Processing, using the credentials specified during catalog creation.

Click Create Catalog in the top-right corner to open the catalog creation page.
Enter a Catalog Name and Description.
Set Catalog Type to External.
Under External Source Type, choose Oracle Autonomous Data Warehouse (you can also select Autonomous Transaction Processing or Oracle Database).
In External Source Method, you can choose either: Wallet – by uploading your ADW Wallet file or ADW Instance – by providing tenant and region details. Here, we’ll use the Wallet method.
a. Upload your ADW Wallet file.
b. From the Service from the drop-down, select a service.
c. Enter your Wallet Password, fUsername, Schema Name, and Password.
d. Click Test Connection to verify that the connection details are correct.
e. Once the test is successful, click Create.
After a few minutes, your new external catalog will be available under Master Catalog.

In the newly created catalog, open the schema that matches your Oracle Autonomous Database user name. You can now browse all tables in that schema and access the data directly from your Oracle AI Data Platform workspace.

Watch the video below for a detailed, step-by-step walkthrough with voice guidance on setting up catalogs with Oracle Autonomous AI Database.

🎥 VIDEO: ▶ AI Data Platform Setting Up Catalog with Oracle Autonomous AI Database

Notebook Management: Export and Import

Notebooks in Oracle AI Data Platform serve as a shared environment for data engineers and scientists to develop code and analyze data interactively. In Oracle AI Data Platform, you can easily upload and download notebooks to manage your work across different environments.

To upload a workbook:

In your workspace, navigate to the folder where you want to store the notebook.

2. Click Upload File.

3. Either drag and drop your notebook file or browse your machine to select it.

4. Click Upload. The notebook will now be added to your workspace.

To down download a workbook:In your workspace, lick the three-dot menu next to the notebook name you want to download.

Click Download to save the notebook on your machine.

This simple process allows you to move notebooks in and out of Oracle AI Data Platform, enabling easy sharing, backup, and offline work.

Watch the video below for a step-by-step walkthrough with voice guidance on uploading, downloading, and managing notebooks in Oracle AI Data Platform.

🎥 VIDEO: ▶ AI Data Platform Notebook Management: Export & Import

End-to-End Medallion Architecture Implementation Example

To wrap up, here’s a practical, end-to-end example of implementing the Medallion Architecture using Oracle AI Data Platform. This linked video demonstrates a Telecom customer churn use case, showing how data moves from raw ingestion to refined, curated layers, and how an AI model is applied to analyze customer sentiments. It highlights best practices for organizing, processing, and analyzing data in Oracle AI Data Platform to generate actionable insights.

🎥 VIDEO: ▶ AI Data Platform End-to-End Medallion Architecture – Telecom Customer Churn