What I appreciated most about DBT was its capability to automate the creation of form data models, allowing me to trust the data. I felt confident that the KPIs displayed were accurate, thanks to transformation logic that had been thoroughly tested and addressed, which I found particularly valuable.

What I like most about dbt is that it brings software engineering best practices to SQL-based data transformations, making our SQL code base maintainable at scale. It has a clear model structure like staging, intermediate, and reporting layers. It provides macros and ref macros that make logic reusable, and the dependencies are really easy to understand. I appreciate its good collaboration with Git and integration with version control. Dbt has a strong documentation background, providing an auto-generated documentation site, so everyone is aware of what's happening in the project. The initial setup of dbt is really easy thanks to its great documentation, and it's available for almost all major data warehouses.

dbt runs well on Redshift, since that is what was mentioned over and over again in the notes; however, dbt simply compiles the SQL and the warehouse itself handles the heavy lifting. Using Git and Version Control for Data Models, is nice because it keeps the data model from exploding. dbt also integrates with our AWS infrastructure without requiring tears. The speed is sufficient, as it simply passes the work to the database; although, having the transformation logic in one location is helpful.