Braintrust uses Codex to turn feature requests into previews
Braintrust says its engineers now use OpenAI’s Codex with GPT-5.5 to turn customer feature requests into preview branches in minutes. The company says half its engineering team moved to Codex in one month, speeding up customer feedback and experimentation.
Braintrust says it is using OpenAI’s Codex with GPT-5.5 to turn customer feature requests into working preview branches in minutes, changing how the company handles product feedback and engineering experiments.
The company, which describes itself as an observability and eval platform for shipping quality AI products, said the workflow lets engineers copy a customer request into Codex, generate a preview branch, and show the result back to the customer quickly. According to founder and CEO Ankur Goyal, that replaces the older process, where requests would go into a backlog and be prioritized later.
“Codex unlocked our ability to try out customer feature requests in real time,” Goyal said in the OpenAI post. He added that the biggest gain is speed, because the tool makes it possible to move from a customer request to a working example without the same amount of manual prompting and back-and-forth.
Braintrust said the faster pace affects more than coding speed. Goyal said it changes the company’s feedback loop with customers, since engineers can now iterate and ideate with customers while the request is still fresh. In practice, that means the company can show a preview branch sooner and use the response to shape the next step.
The post said half of the Braintrust team moved to Codex in one month. Goyal said the speed of the tool matters because it can keep producing output in the terminal without slowing down in the same way as other models, which he said makes it easier to use in real development work.
Goyal also said Codex changed how the team runs experiments. With slower tools, he said, he would have to prompt a model step by step to solve a specific problem. With Codex, he said he can write a test that demonstrates the problem, create a sandbox environment, and let Codex run there.
That setup gives Braintrust a more controlled way to try out ideas. Rather than relying on repeated prompting, the team can define the problem, put the model in an environment built for testing, and move more quickly toward a working solution.
The OpenAI post presents Codex as a tool for turning customer requests into visible prototypes faster, while also giving engineers more room to run experiments inside controlled environments. According to Braintrust, that speed is now part of its development workflow, not just a nice-to-have feature.
Braintrust says the main change is not only that code gets written faster. It is that customer requests can now become preview branches in minutes, with feedback coming back while the conversation is still active.
