← Case studies

Case study 03

Legacy Modernization and Migration

Moved critical workloads off Oracle, SSIS, and MS Access onto Snowflake and automated ETL, cutting an hour-long process to about ten minutes and eliminating 250+ hours of manual work a year.

Snowflake · Oracle · SSIS · Python · MSSQL · DB2 · SQL

~1hr → ~10min
Oracle workload runtime
250+ hrs/yr
manual effort eliminated
~15%
operational cost reduction
Oracle decisioning / ETLSSIS + Python + stored procsMS Access processesAutomated ETLSnowflake + Python
Before
~1 hour
After
~10 min

Oracle decisioning/ETL processing time, before and after the migration to Snowflake.

Context

Not every problem is a greenfield build. Some of the highest-value work is deciding what to rebuild, what to leave alone, and in what order, then actually doing it without breaking what’s running.

Oracle to Snowflake

Critical legacy Oracle-based decisioning and ETL processes were migrated onto Snowflake. Processing time for the affected workloads went from roughly an hour to about ten minutes.

Legacy automation, financial services

At a previous role in financial services, I automated legacy SSIS, Python, and stored-procedure workflows. That eliminated more than 250 hours of manual effort a year and cut operational costs by roughly 15%.

MS Access to automated ETL

Separately, I reengineered legacy MS Access processes into automated ETL packages, integrating sources across Oracle, MSSQL, DB2, Excel, and CSV. Access-based processes tend to be invisible until the one person who maintains them leaves. Turning them into automated, source-controlled ETL removes that risk along with the manual effort.

Judgment

The common thread across these isn’t the technology, it’s sequencing. Migrate the thing causing the most pain first, prove the new pattern works, then extend it, rather than trying to modernize everything at once and risk breaking a process the business depends on today.