Computational Seismologist and Geophysicist for AI Benchmark Design

Design graduate-level computational geophysics challenges that evaluate whether advanced AI systems can perform research-grade scientific workflows using real software.

You will create original problems involving simulations, result interpretation, experiment design, and extracting hidden information from data. The work focuses on rigorous scientific reasoning, not routine data labeling.

• Create computational problems requiring skilled use of specialized scientific software.
• Develop fully specified tasks with exact solutions and more open-ended tasks that require strategic experiment or query planning.
• Design challenges involving what to measure, how to interpret partial results, and how to efficiently narrow possible answers.
• Test problems with advanced AI models and refine them to achieve the intended level of difficulty.
• Write problem setups, oracle functions, and solution validators in Python.

• Graduate-level training in a relevant STEM field, including an MS, PhD, or equivalent research experience. An MS or PhD is preferred.
• Hands-on experience with open-source computational seismology or geophysics tools, such as SPECFEM, ObsPy, Pyrocko, SimPEG, pyGIMLi, SeisBench, EQcorrscan, or Fatiando a Terra.
• Experience with seismic wave propagation, numerical simulation, synthetic seismogram generation, full-waveform inversion, seismic imaging, travel-time tomography, moment tensor inversion, event detection or location, or related computational geophysics workflows.
• Proven proficiency with at least one relevant scientific software library through research publications, open-source contributions, or professional work.
• Strong Python skills and familiarity with scientific code written in Python, C, C++, or Fortran.
• Ability to work independently, incorporate feedback, and identify meaningful edge cases and failure modes in scientific workflows.
• Comfort working in Linux or terminal-based environments with remote compute sandboxes.

• Experience with multiple relevant domains or tools.
• Familiarity with benchmark or evaluation design.
• Experience teaching scientific topics or designing exams and problem sets.
• Experience with computational reproducibility and containerized environments.

• Remote, hourly engagement.
• Availability of at least 15 to 20 hours per week is required.

$70 to $85 per hour.