The work
You will create demanding computational tasks for AI systems and assess whether their technical answers are correct, reproducible, and scientifically sound. Your reviews may cover code, geospatial analyses, simulations, forecasts, and other Earth and environmental science outputs.
You will define clear instructions, reference solutions, and evaluation criteria. You will also check results against observations, measurements, or established models and identify issues with coordinate systems, spatial or temporal data, assumptions, uncertainty, numerical methods, and reproducibility.
- Create realistic computational Earth and environmental science tasks.
- Write task instructions, reproducible environments, reference solutions, and evaluation criteria.
- Review model-generated Python code, geospatial analyses, simulations, forecasts, and technical deliverables.
- Set objective standards for correct technical solutions and assess technical submissions.
- Validate computational results against observations, measurements, or established models.
What it pays and takes
This is a remote, part-time contractor opportunity for candidates with intermediate experience. The schedule is flexible, with a commitment of at least 15 hours per week and a workload of less than 20 hours per week.
- $60 to $120 per hour equivalent, paid per task that meets the project specifications.
- Remote opportunity open to candidates in the listed countries.
- Professional fluency in English.
- Prior paid experience training, evaluating, benchmarking, or improving AI systems on scientific or technical tasks.
- Strong practical proficiency in Python.
- Deep expertise in at least one Earth or environmental science specialization.
- Hands-on experience with geospatial analysis, modeling, simulation, forecasting, remote sensing, statistics, or reproducible research.
- Relevant areas include geoscience, geology, geophysics, atmosphere and climate, meteorology, ocean and marine science, hydrology, remote sensing, environmental science, and sustainability.
- Useful tools include NumPy, SciPy, pandas, xarray, GeoPandas, GDAL, rasterio, and NetCDF, or comparable software.
- Experience creating AI-training tasks, reference answers, rubrics, graders, or benchmark problems is useful.
- Publications, preprints, open-source scientific software, datasets, models, notebooks, maps, simulations, or research profiles can demonstrate relevant work.
Where the role is open
Applications are open to candidates based in the following countries and territories.
- United Arab Emirates, Argentina, Austria, Bangladesh, Belgium, Bahrain, Brazil, Canada, Switzerland, Chile, Colombia, Germany, Denmark, Algeria, Egypt, Spain, France, United Kingdom, Ghana, Greece, Indonesia, India, Italy, Jordan, Japan, Kenya, Kyrgyzstan, Kuwait.
- Kazakhstan, Lebanon, Sri Lanka, Morocco, Mexico, Malaysia, Nigeria, Netherlands, Oman, Peru, Philippines, Pakistan, Palestine, Portugal, Qatar, Saudi Arabia, Singapore, Thailand, Tunisia, Taiwan, United States, Uzbekistan, Vietnam, South Africa.
How it works
You apply here in about two minutes and finish on the hiring platform. The next step is a short AI interview (technical roles include a coding challenge). Passing it is how you qualify for paid tasks, so take it soon after you are invited.
The hiring platform's application for this job asks for your resume, your LinkedIn profile, your phone number, your location and answers to 8 screening questions.
About AI training work
AI training is the human work behind many AI systems, including writing tasks, reviewing model answers, checking code, and rating technical results. People with scientific and programming expertise help set reliable standards so models produce more accurate and reproducible work.