Bart Waardenburg

AI can generate a React component in seconds. The hard part is changing a real React or Next.js codebase safely: knowing what is dead, where logic is duplicated, which boundaries a refactor crosses, and how a change affects the component graph or design system.Coding agents infer this from a prompt and a few open files. That works for local changes, but it misses the repo-wide context that makes larger changes risky. In this talk, Bart shows how agentic engineering changes when Fallow gives the agent deterministic facts about dead code and duplication, architecture boundaries, React hotspots, and design-token drift. The productivity gain comes from giving developers and agents the same map of the system before they edit and a reliable way to verify the result.

AI coding tools are good at writing local diffs, but they still miss repo-wide truth. In large TypeScript and JavaScript codebases, that means dead exports, duplicated logic, accidental boundary violations, and complexity creep after every “small” AI refactor. In this talk, I’ll show a practical workflow for working with large codebases using AI: let the agent generate, run deterministic codebase analysis, feed the findings back via CLI/MCP, and gate drift in CI before it lands. Fallow is the case study, but the workflow applies beyond one tool.