Step is the building block of any ETL. It can be a Source, a Transform, a Destination, or a Visualization.
Visualization step is a special type of Destination, which provides a direct way to see the analysis results.
Sources pull data, Transforms changes data, and Destinations push data out of Desktop application.
To extract any data you'll need to configure a Source step. Sources pull data. Basically each type of source requires different configurations. The UI enables you to explore your datasource by jumping from one page to another with minimum latency.
Once you've got data, you might want to change that data according to your analysis needs. The common scenario is to filter out some records or project only a few columns. If you build metrics, chances are you'll need to aggregate some columns. Usually this can be done with sql transform. Some other scenario requires to join two steps input with a join.
Any Destination Step enables you to store processed data into any supported destination.
Visualization step is a special type of Loader. It saves the data to a temporary and hidden data store, and provide an easy-to-configure visualization interface. You can use Visualization steps to visualize the result directly in Ananas Analytics.