Corey J. Gallon

Gallon’s owned biography describes earlier finance leadership at a Wall Street fintech company, innovation leadership in consulting and teaching graduate machine learning. His conference biography identifies finance analytics work at PwC and teaching at Loyola University Chicago. At Rexmore he applies software and AI systems to building businesses.[1][3]

chrome-agent lets an agent inspect the DOM, accessibility data, screenshots, logs and network activity while sending browser actions through the Chrome DevTools Protocol. Its command-line interface can execute reusable deterministic programs, reducing the need for a model to decide every intermediate step. Gallon’s talk distinguishes a page-level synthetic click from protocol input that follows the browser’s input path; the appropriate mechanism depends on how the page handles events.[4][3]

In The Cure for the Vibe Coding Hangover, Gallon assigns the engineer product intent, interfaces, acceptance criteria and dependency planning. Agents implement small pieces that can be checked independently against persistent specifications. The method aims to preserve a coherent system across many generated changes, rather than accepting an agent’s completion message as evidence that the software works.[5]