Research Engineer, Frontier Evals & Environments - Finance

Job description

About the team The Frontier Evals team builds north star model evaluations to drive progress towards safe AGI/ASI. This team builds ambitious evaluations to measure and steer our models, and creates self-improvement loops to steer our training, safety, and launch decisions. Some of the team's open-sourced evaluations include SWE-bench Verified , MLE-bench , PaperBench , and SWE-Lancer , and the team built and ran frontier evaluations for GPT4o , o1 , o3 , GPT 4.5 , ChatGPT Agent , and GPT5 . If you are interested in feeling firsthand the fast progress of our models, and steering them towards good, this is the team for you. About you: We seek exceptional research engineers that can push the boundaries of our frontier models in the finance domain. We are looking for those who will help shape AI evaluations of financial reasoning and related capabilities, and will own individual threads within this endeavor end-to-end. In this role, you'll: Identify important model capabilities, skills, and behaviors that are crucial to financial workflows, and design methods to quantify performance in these areas Own and pursue a research agenda to identify an important model capability (especially a...