FMCG multi-source data reconciliation for a Fortune-500 NZ subsidiary
Replaced a brittle Excel-driven monthly reconciliation with a scheduled ETL pipeline feeding Power BI. Cut quarter-end from three days to under an hour.
The brief
A Fortune-500 NZ subsidiary in the FMCG space was reconciling sales, distribution and master data across four source systems by hand every month. Quarter-end took three working days and shipped with known gaps.
What I built
- Scheduled ETL in C# with a transactional staging schema in PostgreSQL.
- Validation, exception handling and a real audit trail on every row.
- Power BI semantic model with handoff documentation so the in-house BI team owns it.
Outcome
Quarter-end now runs unattended in under an hour, with errors surfaced before the finance team logs in. Two FTE-weeks per month freed up.