Introduction to the data build tool (dbt)

Cette session sera en anglais/This session will be in English
Introduction to the data build tool (dbt) | Meeting-Join | Microsoft Teams

***

Many people learn from the Jaffle Shop sample dbt project. However, the complexity of the project is quite simple. What I really wanted is a project that goes over the following dbt concepts:

• Seed raw data tables
• Transform raw data using views
• Support multiple databases with macros
• Capture slowly changing dimensions type
• Incremental loading of large data
• Use ephemeral tables as a place holder
• Use dbt utils package to make a date dimension
• Create a dimensional model for reporting
• Save pre-canned queries in analyses directory
• Test data quality using property/schema files
• Output custom data quality exceptions to a table
• Create documentation based on model design

In the end, I created my own sample dbt project using the Chinook database schema as the foundation. I must thank Luis Rocha for the OLTP database design and sample data.
This open source project was built using an Fabric SQL Database as the target. However, it can be built on any supported database using the correct dbt adapter and ODBC driver. At the end of this presentation, a data engineer will be ready to model data for any given database or warehouse.