First step towards a modern data warehouse: Power BI Dataflows

The importance of a data warehouse

In the past, it was necessary to have a solid data warehouse in order to unlock and connect multiple data sources. Both on-premise, as well as cloud-based warehouses, require a lot of manual work in, among other things, installing, feeding, managing and extracting data because this always requires a special ETL (Extract, Transform, Load) tool to make this possible. Even when the data can be easily extracted, it still needs to be visualized. Microsoft has simplified this by adding 'Dataflows' to Power BI.

The benefits of the new Power BI Dataflows

A Power BI dataflow is a process that runs in the background in the cloud. A dataflow process ensures that the (required) data is retrieved and stored in the Cloud (Common Data Service). The Dataflow with different data sources can then be linked in Power BI. When the right relations have been made, the underlying data can be visualised in Power BI.

Working with Dataflows is easy to master and requires little or no knowledge of databases or ETL processes. The advantage we see in working with Dataflows is especially the simplicity of the process. The dataflow can be created in the Power BI browser. When composing our dataflow, we can choose from many common data sources. These include SQL Server, SAP, Oracle, Amazon Redshift, Snowflake, but also Excel, JSON, XML files or an API link. After selecting the desired source, a preview can be used to filter the required data from the system and it can be unlocked, edited, linked and visualized in Power BI desktop.

Besides the user friendliness and simplicity, working with Dataflows is very fast. No long waits until the entire dataset is retrieved, but a preview that allows you to quickly continue. Because of the fact that the data is stored in Azure, the size of the dataset is scalable, which can be useful when more storage is needed.

Only benefits?

The only limitation we have noticed is in the editing of the Dataflow, no entities (data sources) can be merged so Dataflows are strong in unlocking but transforming leaves much to be desired. For this we use the datasets of Power BI. Soon you can expect a blog about the use of Dataflows in Power BI datasets. If you want to get started with Power BI Dataflows, you can find more information here.

Building the bridge between business and IT

Because Power BI Dataflows are fully self-service, the business itself is able to unlock, edit, link and visualise their data easily and quickly within a few clicks. Do you want to discuss how we can support you even more in the area of Business Intelligence? Then get in touch with our BI Consultants.