ETL pipeline migrating legacy banking transaction data to Azure SQL, built with Python, validated in SSMS, and visualized in Power BI. Simulates a real-world core banking migration workflow using production-ready architecture.
| Stage | Details |
|---|---|
| Source | Legacy SAS/Alteryx export with mixed date formats, null values, casing inconsistencies, and typos |
| Transform | Python/Pandas ETL with automated data quality scoring, typo correction, date standardization, and audit trail logging |
| Load | Microsoft Azure SQL Database (serverless) via pyodbc — verified clean in SSMS |
| Visualize | Power BI dashboard connected live to Azure SQL showing transaction volume by region, debit/credit split, and flagged transactions |
| Document | Full enterprise migration doc covering architecture, scope, findings, and recommendations |
- 10 records processed
- 90% data quality pass rate
- 1 flagged wire transfer caught by the pipeline
- Full audit trail on every row
Small dataset. Real architecture. Production-ready patterns.
Python · Pandas · Azure SQL · SSMS · Power BI · pyodbc
| File | Description |
|---|---|
etl_transform.py |
Core ETL pipeline — extract, clean, transform |
load_to_azure.py |
Azure SQL loader via pyodbc |
cleaned_transactions.csv |
Post-transform output |
legacy_transactions.xlsx |
Source data |
Bank_Migration_Data.pbix |
Power BI dashboard |
USBank_Migration_Documentation.doc |
Enterprise migration documentation |
errorsANDcompiles/ |
Screenshots of pipeline execution and SSMS validation |