
Refresh is an AI infrastructure company that builds reinforcement learning environments for coding.

refresh.devCrawled Oct 1, 2026
Benchmarks, evaluations, and datasets for training frontier AI models. Original research on simulation engines for coding, healthcare, and document automation.
An SFT dataset of real developer workflows for training production-grade code models.
An RLVR dataset curated from Gauntlet, delivered in Harbor format for reinforcement learning.
Refresh is hiring ex-founders, GTM engineers, and research engineers building the simulation engines frontier labs train on.
We back confident, self-directed people who see where they can move the needle and go do it. If none of the roles above is an exact fit but you're sure you can help us build the environments frontier labs train on, pitch us the role you'…
Own end-to-end delivery of the strategic projects behind our computer-use and coding data builds. You scope the work with frontier-lab customers, keep it shipping on timelines that matter, and stay hands-on across both the customer conve…
Own the technical surface between Refresh and the frontier labs we sell to. Run evals on prospect models, stand up environments during pilots, write the demos that close deals, and feed what you learn back into product. For people who li…
Refresh builds simulation engines for coding and computer use, for frontier labs to train the next generation of virtual software engineering collaborators.
Refresh builds simulation engines for coding and computer use, for frontier labs to train the next generation of virtual software engineering collaborators.
Refresh privacy policy. How we collect, use, and protect your information.
Refresh terms of service. Rules and guidelines for using our website and services.