How Ramp engineers accelerate code review with Codex

Ramp engineers are leveraging Codex with GPT-5.5 to transform their development processes, primarily by accelerating code reviews. This AI-powered tool provides substantive feedback on pull requests within minutes, a significant improvement over the hours it previously took. The advanced reasoning capabilities of Codex with GPT-5.5 enable it to offer a level of thoroughness that human reviewers often lack the time for, making it an indispensable part of Ramp's code review workflows. Austin Ray, who leads AI DevEx at Ramp, highlights how engineers specifically request Codex, showcasing its value and seamless integration into their daily tasks. This efficiency gain not only enhances developer experience but also boosts overall productivity.

Beyond code review, Ramp is utilizing Codex to develop internal agentic tools, such as the On-Call Assistant. This tool aims to alleviate the burden on engineers during on-call rotations, which are typically complex and demanding due to intricate business logic, extensive domain knowledge, and the need to manage concurrent issues. The On-Call Assistant, supported by Codex's robust reasoning, accelerates development and increases confidence in shipped improvements.

Ray emphasizes the importance of evaluating AI tools based on their tangible impact on shipping code, rather than just impressive demonstrations. For him, the real value lies in how these tools fundamentally change engineering workflows. Codex has proven to be a game-changer for Ramp, allowing engineers to work faster and pursue more ambitious projects.

This shift indicates a new paradigm for engineering, where the role of engineers evolves from writing every line of code to orchestrating AI tools effectively. The key skill becomes knowing how to direct AI, when to trust its outputs, and when to intervene. Ramp's most skilled engineers are those who quickly adapt to this new, AI-augmented approach to software development.