Measuring Coding Agent Readiness: Framework for Enterprise AI Dev – Upsun at dotAI Tech Track AI Day

Why watch

Guillaume Moigneu presents a framework for evaluating AI coding agents in enterprise environments, arguing that agent failures stem from codebase and tooling issues rather than model limitations. The talk introduces spec-driven development and the 8 Pillars of Verification as a methodology to measure agent-readiness before writing prompts. Ideal for enterprise developers, architects, and technical leads looking to implement AI-assisted development with proper verification foundations.