Selected professional work
Reducing manual QA across data releases
I moved a QA automation proof of concept into production, saving approximately eight hours of manual QA per release.
The problem
Monthly Digital CPM and SQAD digital releases required multiple rounds of QA. Newly ingested agency data needed comparison with historical records, earlier releases and broader seasonal or market patterns. This investigation also supported new feature development and testing dashboards.
My role
I worked with multi-million-row pricing datasets using SQL Server/SSMS, Tableau and Power BI. I connected the data investigation with product delivery: requirements, UAT, dependencies and release readiness. I moved the automated QA proof of concept into production.
What the QA covered
The work included new agency data versus historical data, release-over-release comparisons, year-over-year and month-over-month movement, and investigation of seasonal or market anomalies. These comparisons helped put changes in context rather than treating every unusual movement as an error.
The outcome and measurement
The automation saved approximately eight hours of manual QA per release. That estimate was based on manual QA time captured in Jira tickets across multiple validation rounds. It measures reduced manual effort within a recurring delivery process. The scope and measurement basis are specific, rather than an extrapolated annual saving.