# Automation steps

> Add the steps that run after the trigger: notifications, APIs, functions, true/false branches, and record updates.

## Available steps

### [Prompt Rippling AI](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/rippling-ai)

### Notifications

-   Send an email
-   Send a digest email
-   Send an SMS
-   Send a push notification
-   Assign a task
-   [Call a public API](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#call-a-public-api)
-   [Send a Slack message](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#send-a-slack-message)
-   Send a Teams message

### Rippling actions

-   Send a survey
-   Send a report via email
-   [Send a report via SFTP](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#send-a-report-via-sftp)
-   Adjust a time off balance
-   Assign a review
-   Send a document
-   Push a start date
-   Run a function
-   Enroll in a course
-   Enroll in a learning path
-   Make a payment (where available on your payroll experience)

### Data

-   [Query Rippling data](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#query-rippling-data)
-   Refresh a transformation
-   Assign value to a variable

### Custom objects

-   Create records in custom objects
-   Update records in custom objects
-   Delete records in custom objects

### Other

-   Create a Google Calendar event
-   [True/false branch](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#truefalse-branch)
-   [Bulk operations](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#bulk-operations)
-   Delay

You can add multiple steps to an automation. Steps will occur in rapid succession unless you add delays to your automation.

> **Note:**
> The sections below cover the steps developers use most often. The other steps in the list above are covered in [Overview of automation steps](https://help.rippling.com/article?articleId=663137955388) in the Help Center.

* * *

## Using data write steps

Some steps update data in the Rippling database. Examples include create, update, or delete a custom object.

Each data write step is treated as a single unit that either succeeds completely or fails completely. If any part of a step fails to complete, the entire step fails.

For example, if you add a step to update a record in a custom object and there is an error on one of the fields being updated, the entire update will fail, and nothing will be updated in the record.

Once a change is made, it is not reverted if the automation encounters a failure. The changes are committed after each step in the automation. Once a step is successful, its changes are committed and finalized, regardless of whether the automation as a whole succeeds.

* * *

## Using the output of steps in subsequent steps

For some steps, the output of the step is saved within the automation as data that can be used in subsequent steps.

Here are a few examples of how you might use the output of a step in a subsequent step:

-   Send a report to an SFTP server, and then send a notification to your data engineering team that includes the location of the file.
-   Add a step to call an API and use the information from the response to populate a variable in a downstream step.

### Lists as output

Some steps return lists. For example, you can query Rippling data to return a list of people, devices, time cards, etc. When the output of a step is a list, it will only be available to use in a step that completes a bulk operation. Lists cannot be used in other types of steps because the step won’t be able to identify which item from the list to use. For more information, see [Bulk operations](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#bulk-operations).

* * *

## Query Rippling data

You can query the data in Rippling to collect information as input to your automation. For example:

-   Retrieve a list of people who are waiting on work authorization on a regular cadence and send the details to an HR team Slack channel.
-   Look up a list of the people currently on leave and send an email with the details to the person maintaining the shift schedule.

> **Note:**
> This step lets you look up data to use in your automation. If you want to make changes to data using an automation, you can do so by adding a different step. For example, you can create, update, and delete records in a custom object, adjust a time off balance, push a start date, and more.

You can retrieve a maximum of 50,000 records in one automation run. For more information, see [Limits](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/limits).

* * *

## Call a public API

Use the **Call a public API** step to send a request from Rippling to an external API or webhook. The response is saved as automation variables you can then use in later steps.

For more information, see [Call an API or webhook using an automation](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/call-api).

* * *

## Send a report via SFTP

Using automations, you can automatically generate reports and send them via SFTP.

> **Note:**
> The ability to send a report via SFTP requires an Unlimited Advanced Automations subscription.

### Authentication

Each time an automation sends a report to an SFTP server, Rippling needs to authenticate to the server. There are two authentication methods you can use:

-   Password-based authentication
-   SSH key-based authentication

You can select either method in your automation, but SSH key-based authentication is recommended because it provides stronger security than password-based authentication.

To use SSH key-based authentication, you’ll need to generate a public/private key pair, add the public key to the SFTP server, and add the private key to Rippling.

Rippling supports the following encryption standards for decrypting the private key. Other algorithms are not supported: DSA, RSA, ECDSA, and Ed25519. Keys must follow the OpenSSH standard. If your private key is in PuTTY (.ppk) file format, convert the key to the OpenSSH (.pem) file format.

### What is sent

The report will be generated using the automation owner’s permissions. This means that whatever data the automation owner has permission to see will be included in the report. Other collaborators on the automation may not have the same view permissions to the data in the report, so keep this in mind when you’re building your automation.

You cannot use automations to send reports created using one of a handful of special report recipes in Rippling.

By default, reports are dropped in a folder called _/rippling-automations_ on the destination server.

### Use the file path in a later step

When you send a report via SFTP, the full file path of the report, including the file name, is available as output. You can use this in a downstream step to add the full file path to a notification.

You may want to test this step before activating the automation. For more information, see [Automation troubleshooting](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/troubleshooting#send-a-report-via-sftp-fails).

* * *

## Send a Slack message

Use an automation to send a Slack message automatically when something happens in Rippling. After connecting your Slack workspace to Rippling and enabling automation actions, customize the Slack message that is sent:

-   Send a direct message, group message, or a message to a public or private channel.
-   Customize the message that is sent using Markdown to tag users, channels, or special mentions.
-   When sending a message to a public or private channel, choose the destination channel using either a **dropdown** or a **variable**.

### Dropdowns vs. variables

Use a **dropdown** when every automation run should send a message to the exact same channel. The dropdown automatically populates all public and private channels available in your connected workspace via a picklist.

Use a **variable** when the destination channel depends on data from the automation. Instead of creating separate conditions, you can route messages dynamically using a single Slack step.

For example, to use a variable to send PTO requests to a specific channel based on an employee's department:

1.  Create a new [text variable](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/variables) called `DepartmentPTO`.
2.  Add an **Assign value to a variable** step after the automation trigger, and map `DepartmentPTO` dynamically using an [RQL formula](https://help.rippling.com/page?page=rql) that checks the employee's department and returns the matching Slack channel (e.g., `#pto-sales` for Sales or `#pto-engineering` for Engineering).
3.  In the Slack action step under **What channel do you want to send to?**, select **Enter choice using a variable** > **My variables** > click `DepartmentPTO`.

One step can send a message to the correct Slack channel without needing a separate condition for each department.

> **Note:**
> Variable values must be formatted as text matching the exact channel name with a leading `#` (e.g., `#automation-test`). The variable must also already exist in **My variables** to be selected. See [Automation variables](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/variables) for more information.

* * *

## True/false branch

True/false branches let you create a fork in an automation so that it follows one pathway if the condition is true, and a different pathway if it’s false.

You may want to add a true/false branch if you want to use a common trigger but take different steps for different groups, for example, trigger an automation when a new hire accepts their offer, but take one set of steps for remote hires and a different set of steps for everyone else.

When you add a true/false branch, you set up conditions that are evaluated by the automation. If the conditions are met, the automation follows the **true** path. If the conditions are not met, it follows the **false** path. For more information, see [Filter conditions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/filter-conditions).

### When to use a true/false branch vs. conditions on the trigger

-   **Add conditions to the trigger** if you want to narrow down the automation so that it only triggers under specific circumstances.
-   **Add a true/false branch** if you want to expand the automation so that it takes different paths depending on the circumstances.

If you only want the automation to occur for new hires in Sales, you can add the Sales department as a condition on the trigger. The automation will only trigger for people in Sales, and never for anyone else.

If there are also hiring activities you want to automate for people in other departments, you can instead add a true/false branch that checks the new hire’s department, and then takes one series of steps if the person is in Sales and a different set of steps for everyone else.

* * *

## Bulk operations

Use bulk operations to take action on multiple items in bulk within an automation.

For example, you could:

-   Set up a daily automation that finds employees starting in 3 days who have not completed a background check, then assigns a reminder task for each person.
-   Set up a daily automation that calls an API that returns a list of items, then creates a custom object record for each item.

To set up a bulk operation:

1.  Open the automation in Automation Studio and select **+Add a step** > **Bulk operation**.
2.  Optionally name the step and select the type of **loop**:
    -   **Loop a set number of times.** This repeats the step multiple times. Configure the number of times the loop should repeat (using a number or variable).
    -   **Loop over a list.** Use the variable selector to loop over a specific list.
3.  Click **Continue** to save the step.

Each item acted upon in a bulk operation is available as **output** to use in downstream steps. You can navigate to these items using the **Loop** variable.

> **Note:**
> Loop items can include trailing whitespace. In your [function](https://developer.rippling.com/documentation/rippling-platform/developer/functions/getting-started), call `.trim()` on the loop item before comparing it to an API response.

Bulk operations cannot iterate more than 1,000 times. For more information, see [Limits](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/limits).

* * *

## See also

-   [Get started with automations](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/get-started)
-   [Automation triggers](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/triggers)
-   [Filter conditions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/filter-conditions)
-   [Automation variables](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/variables)
-   [Call an API or webhook using an automation](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/call-api)
-   [Automation limits](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/limits)
-   [Automation troubleshooting](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/troubleshooting)
-   [Get started with functions](https://developer.rippling.com/documentation/rippling-platform/developer/functions/getting-started)
-   [Invoke functions](https://developer.rippling.com/documentation/rippling-platform/developer/functions/invoking-functions)
-   [Refresh transformations](https://developer.rippling.com/documentation/rippling-platform/transformations/refresh-schedules)
-   [RQL](https://help.rippling.com/page?page=rql)

Source: https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps

> For the complete documentation index, see [llms.txt](https://developer.rippling.com/llms.txt). Public pages have a Markdown version at the same URL with .md added.
