Have you ever wanted to build a dbt package, but didn’t know where to start? In this 60-minute workshop, Stephen Ebrey, data consultant and founder of Sawtelle Analytics, will break down the process he went through to build the recently-released census_utils dbt package, as well as cover cross DB compatibility, how to use variables to let users customize your package, and integration testing. 

This session will include practical advice on how and why to create a dbt package, as well as notes on challenges of package work that many analytics engineers may not have experience with. In this session, Stephen will cover: 

• Cross DB compatibility
• How to use variables to let users customize your package
• Integration testing

He’ll also give examples of ways the databases differ, such as handling arrays, and the two ways to handle them (dispatch or cross-DB macros), as well as cover which databases have free options and how to switch between dbt adapters to test on them. Stephen will wrap up the session with an example of how a variable can make a macro more useful to a wide variety of users, and then I’ll walk through the basic integration testing process using seed files and dbt_utils.equality.

This live session will be MC’d by Jeff Sloan, product manager @ Census, who helped with QA on the census_utils package development.