Refined Data → Analytics Marts

GCP • Analytics • Intermediate • Retail

Architecture Diagram

Overview

Refined tables are easier to work with than raw data, but analytics teams usually need a more structured model.

Dimensions, facts, KPIs, and business marts make it easier to answer questions about customers, products, sales, returns, and business performance.

In this pipeline, you will build the analytics layer on GCP.

You will use dbt to transform refined BigQuery tables into reusable analytics models, generate KPIs, create business marts, and prepare export views for downstream systems.

What You Will Build

  • Create reusable dimensions and facts using dbt
  • Track customer and product history using snapshots
  • Build customer, product, order, and returns models
  • Generate daily business KPIs
  • Create sales and business marts
  • Build export-ready views for downstream consumers
  • Test and validate analytics models
  • Run dbt transformations using Cloud Run
  • Orchestrate analytics builds using Cloud Composer

Tech Stack

BigQuery • dbt • Cloud Run • Cloud Composer Apache Airflow • SQL

Learning Outcomes

After completing this pipeline, you will be able to:

  1. Build analytics models using dbt
  2. Create dimensions and facts from refined data
  3. Track historical changes using dbt snapshots
  4. Generate business KPIs in BigQuery
  5. Build reusable business marts for analytics
  6. Create export-ready views for downstream consumers
  7. Test and validate analytics models using dbt
  8. Orchestrate dbt workloads using Cloud Composer
View Courses & Pricing →