> For the complete documentation index, see [llms.txt](https://olddocs.exivity.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://olddocs.exivity.io/3.4.3/data-pipelines/workflows.md).

# Workflows

The *Workflows* menu allows you to schedule various tasks and execute them at a specific date and time. This allows the execution of different Extractors and Transformers, so that they are tightly chained together.

## Creating a Workflow

To create a new Workflow, go to '*Data Pipelines*' > '*Workflows*', then click the green button labbeled '*Create Workflow*':

![Creating a Workflow](/files/-Lyn5h-9Dfqkf7Hjh_QQ)

* Provide a meaningful N**ame** for the Workflow
* Optionally you can provide a detailed **Description**
* The **Start date** of the Workflow, when it will run the first time
* The interval: **Hourly**, **Daily** or **Monthly**
* At what time should the Workflow start
* Provide an interval value. I.e. provide 2 for every second  **Hourly** / **Daily** / **Monthly**, depending on your configured interval
* **Add** a new step in your work flow
* Set the **Type** of the step using the drop down menu
* Provide the option that goes with the selected **Type**. This can be your Extractor, Transformer, Report or other name
* Depending on the selected Step Type, you can provide an offset date. This value is used during execution of that step. Typically this would be used for a **From** date offset (i.e. **-1** for yesterday)
* A **To** date offset can be provided for some step types (i.e. **0** for today)
* In case multiple steps need to be executed in parallel, you may chose to remove the **Wait** checkbox. 
* Additional arguments can be send to the step. This applies only to some **Extractors** and when executing a custom **Command**
* You can delete a step using the red minus button
* To view historical Workflow results, click the **Status** tab

### Special Workflow Steps

Apart from adding *Extractor*, *Transformer* and *Report* steps, there are two different *Workflow* Step types:

* Core
* Execute Command

#### Core

The *Core* command allows you to run a few predefined API calls, currently these are the following:

![Core API calls using workflow step](/files/-LNPrWL8Oh3tUQA67IZK)

* *Run garbage collector*
  * Cleans up the server getPrices cache table and Redis cache
* *Purge cache*
  * This will Unprepare any prepared reports. Use with caution
* *Refresh budgets*
  * Evalualte all configured budgets
* *Send heartbeat*
  * Send an API heartbeat request. For future use.

{% hint style="danger" %}
**Purge cache** should be used with caution, as it will unprepare **all** available reports. This means that nonr of your reports will return any data, until you have prepared them again.
{% endhint %}

#### Execute command

The Execute Command step enables you to execute an external command, like a script:

![Call external commands or scripts](/files/-LNPtR2AV9DTUYWeRoNY)

As an example: you could run a Powershell script to obtains some data from a special data source that Exivity Extractors are not able or allowed to connect to. This script could be executed the following manner:

```
powershell.exe "D:\script\special.ps1" ((get-date).addDays(-1)).ToString("""yyyyMMdd""")
```

The above command calls the Powershell executable to run the *special.ps1* script, with a dynamically generated parameter that is evaluated at run time. This particular example always provides yesterdays date in *yyyyMMdd* format as a parameter to the *special.ps1* script. Many other variations and scripting languages are possible. Feel free to experiment.