Methodology
I hold a B.S. in Electrical Engineering and completed a Full-Stack Software Engineering immersive before moving into QA. This deep technical background is why I can read a codebase and understand the systems underneath. The knowledge of how an application works is the foundation I use to design stable test infrastructure and meaningful coverage. AI tooling extends the scope and speed at which I can work.
The method
-
Learn
I learn how an application works before testing it. I look at the UI, the database, the API calls, how data flows, and any external systems.
Bank of America · 2024–2026
At BofA the stack was Angular, .NET, and Kafka. A story was not finished until I had verified complex business logic against the UI, database, and Kafka topic messages.
I knew the stack well enough to pick up small feature stories alongside my QA work.
-
Build
I build test infrastructure that doesn't affect real-world systems by mocking data and decoupling dependencies.
Sovrn · 2022–2024
At Sovrn, I built Playwright infrastructure in TypeScript that simulated ad slots on publisher sites so that auctions could be tested without affecting Google Ad Manager.
-
Gate
I wire test runs into CI so that quality can be enforced with minimal friction.
Sovrn · 2022–2024
I tooled the full CI pipeline using GitHub Actions, Webpack coverage instrumentation, and Datadog reporting.
I simplified setup and seeding of the test framework to one command, enabling developers to run the tests locally.
-
Set
All framework decisions, reference tests, conventions, and UX patterns get codified into AI context.
Bank of America · 2024–2026
I wrote context and rules files for Copilot to generate manual test cases from acceptance criteria alone.
This eliminated 80-90% of the repetivie and tedious work allowing me to focus on automation and edge cases instead of writing audit trails.
-
Scale
AI applies my work across the suite where I review every output against the original story or feature it was meant to follow.
Bank of America · 2024–2026
GitHub Copilot ported the remaining Playwright specs to the reference pattern, and drafted roughly 90% of a manual test case from a story's Gherkin acceptance criteria.