Teams at Epoch AI

Build and maintain cybersecurity benchmarks for evaluating frontier AI models, primarily using the Inspect library. Secure and harden evaluation sandboxes against capable adversaries, apply threat modeling to benchmarking strategy, integrate AI providers, and maintain reliable evaluation infrastructure. Collaborate with researchers, analysts, and engineers to ensure accurate results and support public research products. The role also involves proposing new experiments and staying current with AI agents, coding harnesses, and cybersecurity red teaming.

Own the administrative and data operations supporting hiring, including ATS accuracy, hiring-round setup, reporting, job postings, work-test administration, candidate communications, interview scheduling, and recruiter support. The role also provides opportunities to assist with application review, sourcing, and broader People Operations. Success requires strong organization, attention to detail, clear written communication, spreadsheet and reporting skills, discretion, and comfort working remotely across time zones.

Design responsive interfaces, dashboards, graphs, and data visualizations that communicate complex AI research clearly. Create specifications, mockups, and prototypes; maintain design systems and visual consistency; and collaborate with researchers, engineers, and communications staff to deliver polished products. Depending on specialization, the role focuses on product and web design or data visualization. The position is fully remote with flexible hours and attendance at three in-person retreats annually.