This role involves turning mathematical and research workflows into terminal-based tasks to test and evaluate AI agents across numerical analysis, optimization, statistics, and scientific computation. Candidates must have advanced technical experience (such as a Ph.D. or postdoctoral work) in mathematics, statistics, or related fields, alongside proficiency in languages like Python, R, Julia, C/C++, or Bash. The position is a remote, worldwide, 5-week contractor role requiring 40 hours per week with at least 4 hours overlapping with Pacific Standard Time.