Software Engineer, Benchmarking

• Help evaluate frontier AI models by developing and maintaining benchmarking infrastructure
• Implement AI benchmarks within evaluation infrastructure to expand tracked capabilities
• Develop existing benchmarks to quickly evaluate new model releases
• Contribute to brand new benchmarks by pitching and prototyping your own ideas

This role contributes to the evaluation of frontier AI models, helping to advance understanding and development of AI capabilities through systematic benchmarking.

• Software engineering skills for developing and maintaining benchmarking infrastructure
• Experience with AI model evaluation and benchmarking
• Ability to collaborate with researchers and analysts
• Creative thinking for pitching and prototyping new benchmark ideas

No benefits information provided in the job description.

Epoch AI appears to be an organization focused on AI research and evaluation, specifically working on benchmarking frontier AI models.