Faster Coding Requires Better Code

Many codebases already struggle with technical debt that makes development expensive, frustrating, and risky. When coding agents enter the development loop, existing technical debt becomes an AI-adoption constraint. Poor-quality code causes more defects and turns even state-of-the-art AI agents into legacy code generators.

In this keynote, Adam Tornhill shows why code health cannot be treated as a mere engineering concern: it determines how much of the promised AI productivity gains organizations can capture. Backed by large-scale empirical studies on AI coding and developer productivity, we separate what works from what doesn't in real-world systems. Building on these findings, we distill practical techniques for scaling agentic coding while preventing a quality crash.

About Adam Tornhill

Adam Tornhill is a programmer who combines degrees in engineering and psychology. He’s the founder of CodeScene where he designs tools for code analysis. Adam is also the author of multiple technical books, including the best selling Your Code as a Crime Scene and Software Design X-Rays. Adam’s other interests include modern history, music, retro computing, and martial arts.

More About Adam »