Before you create your pipeline, determine the following:
What data you need for your workflow
Determine what information you need to access to complete your process.
Which software services you need to include
Based on where this data is located, decide which cloud-based software services your want to use. In Pipelines, these software products are referred to as channels.
How you want your pipeline to initiate:
Your pipeline can start one of following ways:
- On a schedule. You can choose the exact time and/or frequency to run your pipeline
- Manually, by clicking Run pipeline at the top of the page. Running a pipeline manually is particularly useful for one-time data imports. To run a pipeline manually, your first step must be either an action or query step.
- Trigger based. You pipeline can initiate based on one of the following:
- An external webhook
- A data change in a channel
- From another pipeline
Your workflow may involve multiple pipelines. They can be organized in a combination of Trigger based and scheduled or one-time run pipelines. A pipeline can call another pipeline.
Whether you need to add filters to your steps
You can specify criteria that narrows the data your step returns.
Whether you need to add conditions
Your pipeline can include conditions, such as if-then statements, for-each loops, and stop blocks.
Whether you need to transform your data
Whether you need more than one pipeline for your workflow
Each pipeline can include only one trigger. If you have a complex workflow you may need multiple pipelines to perform all actions and trigger on all events needed.
For example, a workflow to sync information between Jira and Zendesk could include multiple pipelines.