Summary of Mastering the art of dbt package development

00:00:00 - 00:20:00

This YouTube video discusses the various aspects of developing dbt packages, including how to select tables and fields, how to use macros, and how to perform grain and null testing. The presenter also discusses the importance of engaging the community in package development.

• 00:00:00 This video discusses the three main components of a recipe for developing DBT packages, how the DBT package development team gathers ingredients, how they develop and test the packages, and how they engage the community.
• 00:05:00 The presenter discusses the philosophy and approach they take when developing DBT packages. They explain that they aim to create models that can get customers 80% of the way to their reporting needs, while providing flexibility and opportunity to customize their own projects. They also discuss the use of tables and fields, and how they select which tables and fields to include in a package based on the usage of those tables and fields.
• 00:10:00 The speaker discusses various aspects of dbt package development, including how variables and fields are taken into account, how discovery sessions help identify additional requirements, and how flexibility and customization are offered. They mention a style guide for SQL and YAML, and mention a plan to develop a playbook for different package types. They also mention the use of macros, and discuss how their team tries to adhere to consistent coding practices.
• 00:15:00 The speaker discusses the different models that DBT packages can take, including source (DBT platform suffix with Source), modeling (DBT platform prefix with int), and staging (CR). They mention that Stage 1 models are meant to be as close to the source data as possible, and that Stage 2 models are used to obstruct complexity from the end user. Intermediate models are used to reduce complexity from the end user, and final models are used to provide a quick glance understanding of the final resulting model. Integration tests are important for quality assurance, and standards for release are to pass integration tests before merging any PRs into Main.
• 00:20:00 The speaker discusses the basics of package development and how two tests, grain testing and null testing, can help to ensure accuracy and consistency of models. They also mention that Community contributors are always appreciated.