Olist E-commerse

This project is an end-to-end data engineering pipeline built to analyze Brazilian e-commerce trends using the Olist public dataset, covering approximately 100,000 orders placed between 2016 and 2018. The pipeline follows a medallion architecture ingesting raw CSV files into AWS S3, creating external Athena tables, and using Bruin to orchestrate staging transformations (type casting, null handling, data quality checks) and mart aggregations, all materialized as Iceberg tables. The final mart layer feeds a Looker Studio dashboard answering three core business questions: how daily revenue evolved over time, which Brazilian states drive the most e-commerce value, and which product categories are the top revenue generators. The goal of the project is to demonstrate how Bruin can replace a traditional multi-tool stack consolidating orchestration, transformation, and data quality into a single framework while delivering production-grade pipeline outputs backed by 24 automated quality checks across 15 assets.