A seasoned testing consultant shares two real-world productivity improvement stories. The first (4x) examines how a tester working at 50% capacity delivered double the output compared to full-time work a year earlier, attributed to accumulated domain knowledge, risk-based testing, and specification involvement. The second (13x) compares two phases of test automation work, where GitHub Copilot agentic use was introduced in the second phase. The author concludes the dramatic improvement came primarily from learning and accumulated context rather than the AI tool itself, while noting risks like drift in quality under pressure.
Sort: