Data Integration in Einstein Analytics
This Blog explains various ways to integrate and load data into Einstein Analytics
Data Integration is one of key aspects of BI Tool, Einstein Analytics does exceedingly well in this department with seamless integration, as it doesn't require data to in particular format (like star or snowflakes schema).
Unlike the other BI tools, Einstein Analytics stores the data itself in the cloud not just the metadata. Hence, we need to refresh data time to time. In order to ensure we are working or using the latest data. Storing the data along with inverted index (the way data is stored in Einstein Analytics) boost the performance of the tool.
The options available to load data into Einstein Analytics are
- Salesforce: The salesforce objects can be directly loaded into Einstein Analytics using dataset builder or dataflow.
- CSV: Csv files can be directly uploaded into Einstein Analytics.
- Informatica Rev: The data from external data source that can be loaded into Einstein Analytics using Informatica Rev.
- ETL Connector: ETL tools like Informatica cloud, mulesoft, Boomi, snaplogic, etc have connector for Einstein analytics using which you can load data into Einstein analytics.
Screenshot taken from Einstein Org
The data integration can be done at
Dataflow- Dataflow view in Einstein Analytics has etl transformation like Extract, Augment(Join), Append(Union), Slicer, etc., One can leverage this option to integrate data coming from different data sources. Dataflow is also used for data refresh in Einstein Analytics.
Using ETL Connectors- (Eg: Informatica Cloud, the data can be integrated at informatica cloud and integrated data can be loaded into Einstein analytics using "salesforce analytics" connectors.
- Binding: The changes in one step/widget triggers change in other step/widget in a dashboard, this is achieved using binding if the step(s)/widget(s) are created using 2 different datasets.
- Connect Data Source: This option is available at dashboard level, using which you can connected two columns from two different datasets.
- SAQL: You can write a SAQL query to fetch data from two different datasets at dashboard level.