Automate Data Studio Workflows with Jobs in Autonomous AI Database

Use the Jobs feature in Data Studio to automate tasks that you run from Data Studio in Oracle Autonomous AI Database. A job can include steps such as running a Data Load task, running a Table AI Assist recipe, running a PL/SQL procedure, or making a REST call. Jobs can be scheduled to run at defined intervals or run manually when needed.

Use this Autonomous AI Database topic only as the entry point for understanding when and how to use Jobs with Autonomous AI Database. See Data Studio Documentation for complete reference about Jobs.

When to use Jobs

Use Jobs when you want to automate a repeatable Data Studio process instead of running each task manually.

For example, use Jobs to:

Run a saved Data Load task.
Run a saved Table AI Assist recipe.
Run a PL/SQL procedure in the database.
Call an external or cloud REST endpoint.
Combine multiple steps into one workflow.
Schedule the workflow to run at a selected interval.
Run the workflow manually from Data Studio, REST APIs, or SQL.
Monitor each job run and review row counts, rejected rows, errors, and execution details.

Use Jobs when the workflow belongs inside Autonomous AI Database and can be represented as Data Studio tasks, PL/SQL procedures, or REST calls. Use another scheduler or orchestration tool when the process must coordinate services outside the scope of Data Studio or requires broader enterprise orchestration.

How Jobs work with Autonomous AI Database

Jobs run from Database Actions for Autonomous AI Database. Database Actions is bundled with each Autonomous AI Database instance and provides development tools, data tools, administration, monitoring, and download features. Data Studio is one of the Database Actions feature areas. See Connect with Built-In Oracle Database Actions for more details.

A job is made up of one or more steps. Each step performs a specific action, such as a data load, Table AI Assist recipe, PL/SQL procedure, or REST call. When the job runs, Data Studio executes the steps in the order defined in the job. You can select whether a step stops the job on error or allows the job to continue. You can create a job with general job details, add steps, configure reporting notifications, define a schedule, and review API details for running or checking the job from REST or SQL. See Create a Job for more details.

Note: Each time a job is run, the full flow starts again from the beginning. Design job steps so that repeated runs do not create unintended duplicate data, overwrite the wrong target, or repeat an external action unexpectedly.

Before you begin

Before creating a job, prepare the tasks that the job will automate.

For example:

Create the Data Load task that the job will run.
Create the Table AI Assist recipe that the job will run.
Create the PL/SQL procedure that the job will call.
Gather the method, URL, headers, body, and credential details for any REST call step. See How to Access the Jobs Feature for more details.

You also need the appropriate database scheduler privileges. To create jobs in your own schema, you need the CREATE JOB privilege. To create jobs in another schema, you need the CREATE ANY JOB privilege.

If you want job notifications by email or Slack, configure those settings in Data Studio Settings before using them in the job. SMTP settings are used for email notifications, and the Slack tab is used to configure Slack credentials and a channel ID for Jobs notifications. See Data Studio Settings for details.

Open Jobs

To open Jobs:

Open Database Actions for your Autonomous AI Database.
Select Data Studio.
Select Jobs.
Use Create Job to define the job name, steps, reporting options, schedule, and API details.

What you can do with Jobs

Task	Description
Create a job	Define the job name and optional description, then add one or more steps. See Create a Job for guidance.
Add a Data Load step	Run a previously created Data Load task as part of the job. See Create a Job for guidance.
Add a Table AI Assist recipe step	Run a previously created Table AI Assist recipe as part of the job. See Create a Job for guidance.
Add a PL/SQL procedure step	Run a database procedure and pass arguments to it. See Create a Job for guidance.
Add a REST call step	Call a REST endpoint using a selected HTTP method, URL, credential, headers, and body where applicable. See Create a Job for guidance.
Configure notifications	Send email or Slack notifications for job events such as loaded rows, failures, exceeded runtime, or rejected rows. See Create a Job for guidance.
Schedule a job	Enable scheduling and define the interval, weekdays, start date, start time, end date, and end time. See Create a Job for guidance.
Run a job manually	Run a job from the Jobs page, from REST API details, or from SQL shown in the API Details tab. See Create a Job for guidance.
Manage a job	Delete a job, run it immediately, enable scheduling, or disable scheduling. See Manage and Monitor Job for guidance.
Monitor a job	Use the Report feature to review job execution details, including log date, action, step type, step number, rows loaded, rows rejected, error message, and details. See Manage and Monitor Job for guidance.
Automate with PL/SQL	Use the `DBMS_DATA_TOOLS_JOB` package to create, update, rename, run, stop, or delete jobs programmatically. See DBMS_DATA_TOOLS_JOBS Package Reference for reference. Jobs are created disabled by default when using the DBMS_DATA_TOOLS_JOB.CREATE_JOB procedure unless the enabled attribute is set to TRUE. A disabled job exists as a database object, but the scheduler does not run it until it is enabled.