AI Trainer (Physical Sciences)

What You'll Do6

• Design realistic computational physical-science tasks that test advanced AI models.
• Write task instructions, set up reproducible environments, and define reference solutions and evaluation criteria.
• Assess model-generated calculations, code, simulations, and technical deliverables.
• Spot invalid assumptions, numerical errors, physically impossible results, incorrect methods, or reproducibility failures.
• Create and maintain benchmarks or rubrics for judging solution correctness.
• Document evaluation outcomes and provide clear feedback on model performance.

Requirements5

• Prior paid experience training, evaluating, benchmarking, or improving AI systems on scientific or technical tasks.
• Strong practical proficiency with Python.
• Deep expertise in at least one physical-science specialization such as physics, astronomy, cosmology, materials science, chemistry, computational physics, computational chemistry, or scientific simulation.
• Hands-on experience with computational methods for modeling, simulation, numerical analysis, data analysis, or reproducible research.
• Public evidence of skill, such as peer-reviewed publications or preprints involving computational methods, public simulation or analysis code, Google Scholar/ORCID/arXiv profiles, open-source scientific software contributions, reproducible notebooks/models/datasets/simulations, or public benchmarks/experiments/technical projects.

Preferred Qualifications4

• Experience creating AI-training tasks, reference answers, rubrics, graders, or benchmark problems.
• Familiarity with NumPy, SciPy, SymPy, Astropy, PyTorch, JAX, or similar scientific tools.
• Experience validating simulations or models against analytical results, observations, or experimental data.
• Availability to begin immediately.

Who Should Apply

Computational physical scientists with paid AI-training experience who can design and evaluate technical tasks. You should have strong Python skills, deep expertise in a physical-science domain, and public evidence of your computational work.

Salary Insight