[Your Name] · [Email] · [Phone] · [City, ST]
April 21, 2026
Dear Hiring Manager,
I'm applying for the Senior SDET role on your Checkout team. Your RFC on moving from a nightly Cypress run to sharded Playwright on every PR — which I saw on your public engineering discussions — is almost the exact migration I led at HubSpot last year, and the places where you flagged flake risk are the same places I hit.
At HubSpot I owned the test automation strategy for the Marketing Hub platform. I replaced a 3,800-case Selenium suite that took 4h 20m and flaked on 14% of runs with a Playwright + pytest suite that runs 2,400 high-signal cases on every PR in 7 minutes with a 0.6% flake rate. We got there by deleting 1,400 duplicate or low-value cases rather than porting them — a call that took three months of stakeholder work but was the single highest-leverage thing I did that year. Production escape rate dropped from 2.3 customer-reported regressions per release to 0.4, and the release cadence moved from every two weeks to twice weekly. I wrote the flake-diagnosis runbook that's now required reading for anyone added to the test-platform on-call.
Before HubSpot I spent two years as the second QA hire at a Series B healthcare startup (Sana) where I built the first automated test suite from scratch, wrote the performance test harness in k6, and owned the HIPAA evidence collection for every release. That compliance-adjacent work is why I care about traceability — being able to point at a test run and say which control it satisfied is underrated in most QA shops. Your team's decision to treat test data as a first-class product (I saw your data-fixture generator open-sourced last fall) is the kind of investment that tells me you take quality seriously.
I'd love to walk you through the Playwright migration and hear where your team is on the flake-budget conversation. I can share a redacted copy of our flake runbook or jump on a call whenever works.
Sincerely,
[Your Name]