Google Meet is preparing a 10-week rollout of a new live caption reliability upgrade across Android, iOS, and web. You are the QA Engineer responsible for presenting your testing methodology as an execution plan, not just a list of test types. The core team includes 6 software engineers, 2 QA engineers, 1 product manager, 1 UX researcher, and 1 site reliability engineer. Leadership wants the feature launched before a major Workspace customer event, so the timeline is fixed.
The Google Meet PM wants broad launch coverage and a simple go/no-go recommendation. Engineering wants to minimize late-cycle test churn because two engineers are also supporting production incidents. The SRE lead cares most about rollback readiness and production monitoring. The Accessibility lead wants strong coverage for accented English and noisy environments, even if that reduces scope elsewhere.
You have 10 weeks total, a remaining QA tooling and vendor budget of $85,000, and no additional headcount. The feature must support 3 surfaces (Android, iOS, web), 5 launch locales, and 12 critical device/browser combinations. A dependency on the speech model team will not be finalized until the end of Week 4. Only 1 week is available for full end-to-end regression before launch.