How Ramp Engineers Speed Up Code Reviews with Codex | OpenAI

May 20, 2026

How Ramp Engineers Speed Up Code Reviews with Codex

The team uses Codex, powered by GPT‑5.5, to review code and build agents for managing on-call rotations, improving developer experience and boosting productivity.

Image 1: Ramp customer story artwork for Codex.

Company size: Enterprise

Region: North America

Industry: Technology

Product: Codex

At Ramp, engineers are using Codex, powered by GPT‑5.5, to accelerate code reviews and building internal agentic tools that help the team get substantive pull request feedback in minutes instead of hours. Thanks to its reasoning capabilities, Codex with GPT‑5.5 can uniquely reduce a lot of the manual, human work they would otherwise need to do.

“Codex code review catches things that I miss, things that other engineers miss, and of course things that other AI code review tools would definitely miss.”

—Austin Ray, Ramp AI DevEx

Running code reviews teams can trust

Ramp’s AI Developer Experience (AI DevEx) team is using Codex to improve software development speed and code quality.

“Codex code review is the gold standard in the industry. We’ve been leaning on it at Ramp for a long time,” explained Austin Ray, who oversees AI DevEx. “It’s really good, and our engineers call for it by name. People are eager for it to comment on every PR, and it’s become a necessary part of many code review processes.”

Previously, Ramp engineers often waited hours for the first round of review; now, they can get substantive feedback from Codex in minutes. What makes Codex different is that it reasons deeply across the entire codebase, bringing what Ray described as “the level of rigor that most human reviewers don’t have time for.”

Codex combines that depth with what Ray called “an experience that matches how engineers work today.” Engineers who prefer to work close to the metal can use the CLI, while the Codex app provides visual prompts, utilities, and more for those who need them. Ray usually works in the CLI, but he was drawn to the app too. “It has this feeling of an app that is going to steer you toward more productivity in your engineering workflow,” Ray said.

Video 1

00:00

“Codex, powered by GPT-5.5, is incredibly adept at handling that complexity, and it would probably take me a lot of brainpower, a lot of sleep, and a lot of focus on the problem to figure it out myself.”

—Austin Ray, Ramp AI DevEx

Building internal tools with Codex

Ray is also using Codex to support the development of On-Call Assistant, an agentic tool that can take on much of the workload for Ramp engineers during on-call rotations.

“On-call is hard,” Ray explained. “We have a lot of business logic, domain knowledge, and major incidents. You have to hold a lot in context and reason through complicated problems.”

That’s not easy for engineers. It takes a lot of brainpower, and even more sustained, uninterrupted focus.

“There’s just so much complexity,” Ray said. “There are lots of concurrent bugs, there’s a tricky balance between external and internal events, and there are long-running incident investigations where details keep changing and you have to stay on top of them.”

With Codex, Ray can rely on its “extremely strong” reasoning to support development. As a result, On-Call Assistant is being built much faster, and Ray has more confidence in each improvement as it ships.

“Our product surface area is huge,” Ray said. “Codex with GPT‑5.5 just chews through it.”

Leadership lessons

First, Ray is a platform engineer, and he evaluates all developer tools, including AI-powered ones, from that perspective. As he puts it: “Does it really change how people deliver code, or is it just a demo?”

That’s also the advice Ray gives other leaders: focus on real hands-on experience and real outcomes.

Demonstrate the potential of AI tools firsthand: “Have your engineers install Codex, sit with them, and walk them through a solid first experience. Paint a picture of what the future of development can look like.”
Build a path to trust and iteration: “Most engineers don’t fully understand or trust that they’re going to have a good experience with this. They think of it as an experimental tool. By guiding them through that first experience, you change their perspective and get them to keep exploring and iterating on their own until they become one of your best AI users.”
Invest in the feedback loop: “We talk directly to the Codex team about feedback. When issues come up, we have a direct line of communication. That feedback loop is what makes the vendor relationship worth investing in, and we’ve made amazing progress with the Codex team.”

“Codex is the real deal. Codex has genuinely helped us ship faster.”

—Austin Ray, Ramp AI DevEx

What’s next

Codex is changing how quickly Ramp engineers work and giving them the resources to support bigger goals. For Ray, that signals a broader need to think about engineering in a new way.

“Engineers are going to become orchestrators. The skill is no longer writing every line of code yourself; it’s knowing how to direct AI tools like Codex, when to trust them, and when to push back. At Ramp, our best engineers are the ones who are learning that fastest.”