Woofun AI reports that Ryan Lopopolo has joined Google Cloud as the Agentic Google Cloud Platform Chief Engineer, transitioning from his role at OpenAI. At OpenAI, Lopopolo authored the "Harness Engineering" methodology, which focuses on configuring agents with context, tools, permissions, evaluation, and feedback loops for active software development participation.

Lopopolo previously led an internal OpenAI experiment utilizing Codex to build a product from an empty repository, generating approximately 1 million lines of code over five months. In this process, Codex handled code generation, testing, CI, documentation, and internal tools, while humans focused primarily on goals and acceptance criteria. This appointment signals Google Cloud's strategic shift toward enhancing enterprise agent engineering capabilities alongside model and compute offerings.