This repository displays some examples of the transformation pipeline that I have built and maintain at Truvani. No private or proprietary information is included, and all table names and filter conditions have been generalized.
The model directories in this repository roughly follow dbt's conventions. staging
models connect to the data sources (usually Airbyte) and handle the most basic transformations that are needed for the rest of the pipeline. intermediate
models then perform intermediary calculations, aggregations, and joining of multiple data sources. Most of the work is done here. Finally, marts
models perform the very final calculations needed and are accessed by our dashboarding platform, Metabase, to display these metrics to analysts and stakeholders. These marts
models are organized by the team for whom they are designed: dtc
models make up the bulk of the project and provide KPIs for our DTC marketing team, finance
models aggregate metrics for our finance analysts' reporting, and retail
models display information about our products' performance in retail stores across the US.