Use the Retrieve Information action to extract data from a model or perform a real-time search across your content, then use that data later in your workflow. This makes your workflows more innovative and more dynamic by pulling in up-to-date information to personalize actions, guide decision-making, or feed into AI-generated messages.
When to use
Use this action when you want to:
- Look up related details from a model, such as a manager, department, or employee status.
- Search knowledge or document models using keywords or properties and pass the results into later actions.
- Temporarily store retrieved data so you can reuse it in follow-up steps—like assigning tasks, triggering conditions, or generating text.
- Personalize workflows based on the user's context or the latest information in your system.
Global search option
You can choose between two sources when configuring this action:
- Model: Look up data directly from a specific model in your system.
- Global search: Perform a keyword search across one or more models, such as Knowledge or Documents, and retrieve relevant results. This is ideal for surfacing content like policies, guides, or templates that match your search criteria.
You can also:
- Restrict the search to specific models or properties.
- Choose the user context to run the search as (for example, the triggering user) to ensure permissions are respected.
- Limit the number of results returned.
- Access the retrieved record count using a token (e.g.
retrievedKnowledgeResults.retrievedRecordsCount) to control flow logic.
Example: Send approval to an employee's manager
To send an approval request to the employee’s manager:
- Add a Retrieve Information action to fetch the manager’s name and email from the user's record.
- Add a Send Notification or Assign Task action and insert the retrieved data to direct the task to the manager.
Example: Generate AI content using knowledge articles
To generate a welcome message for new starters based on the latest HR policies:
- Add a Retrieve Information action and choose Global search as the source.
- Search your Knowledge model using the triggering user’s job title or department.
- Add a Generate Text action and insert the retrieved results into your AI prompt using tokens like
Retrieved Knowledge results.
This ensures your AI-generated message reflects the most relevant and up-to-date content in your platform.