
Quality-First Agentic Workflows in Practice

Why Your AI Strategy Is Failing: The AI Paradox of Optimizing Coding Alone - Andrew Haschka
Topic 5 of 19
Published May 18, 2026 · Clip 14:33–18:53 (4:20)
Andrew shares customer examples: Deakin University aiming to improve developer experience while consolidating toolchain and quality, and a Thai entertainment/retail CTO delivering hourly with under 2% failures who sees quality as the main AI opportunity. He stresses that quality requirements are built in the planning phase—"rubbish in, rubbish out"—and that thinking holistically about quality outcomes matters more than traditional speed and stability metrics. An agentic flow should curate well-documented requirements, adequate test coverage, compliance checks, and pre-release validation.