Building self-improving tax agents with Codex (17 minute read) (opens in new tab)

Real-world systems often behave differently in production than they do in the lab. Teams often discover these failures after launch, then spend weeks fixing them. That feedback loop is slow and manual. Today, it is possible to build agents that self-improve. This post looks at how OpenAI used Codex to build this type of agent at Thrive Holdings, resulting in an AI that can prepare increasingly complex tax returns.