fash112/nigeria-automotive-data-platform

End-to-end automotive dealership data platform — medallion architecture (bronze/silver/gold), dbt transformations with tests, DuckDB # Nigeria Automotive Sales & Service Data Platform An end-to-end data platform for an automotive dealership and repair business — ingestion, transformation, testing, orchestration and BI delivery. Built on the medallion architecture with Azure and dbt. **Author:** Dayo Fasokun — Data & Infrastructure Engineer **Stack:** Azure Data Factory · Azure Data Lake Storage · dbt · Python · PySpark · SQL · Power BI **Data:** 100% synthetic. No real customer, vehicle or financial data is used anywhere in this repository. --- ## Why this project exists Early in my career I worked as an administrator at an automotive company in Lagos. Work orders lived in paper files and spreadsheets. Parts inventory was counted by hand. Nobody could answer basic questions: which mechanics were most productive, which repairs were actually profitable, which customers were about to stop coming back. This platform is the data infrastructure that business needed — rebuilt properly, seventeen years later, with the tools I use today. --- ## The business questions it answers | # | Question | Served by | Status | |---|----------|-----------|--------| | 1 | Which service types generate the most gross profit after parts and labour? | `mart_service_profitability` | Built | | 2 | Which mechanics complete work fastest without triggering rework? | `mart_technician_performance` | Planned | | 3 | Which parts will stock out in the next 14 days? | `mart_inventory_health` | Planned | | 4 | Which customers are overdue for service and likely to churn? | `mart_customer_retention` | Planned | | 5 | What is the true cost of a repair job, end to end? | `mart_job_costing` | Planned (`fct_work_orders` already computes per-job cost and profit) | | 6 | Which vehicle models generate the most repeat repairs? | `mart_vehicle_reliability` | Planned | --- ## Architecture ``` SOURCES BRONZE SILVER GOLD ───────── ────── ────── ──── Work …