Easy JobAI Trainer (Earth & Environmental Sciences)
micro1 · Worldwide · Contract · Remote
Pay: $60–$120/hour equivalent, paid per accepted task
Location: Global, fully remote
Job Type: Contractor
Time Commitment: 15+ hours per week
Schedule: Flexible
micro1 is looking for Computational Earth and Environmental Scientists with prior professional AI-training or model-evaluation experience to support a project focused on improving advanced AI systems.
You will use your scientific expertise and programming skills to create challenging technical tasks, evaluate model-generated code and analyses, identify failures, and establish objective standards for correct solutions. This role is intended for candidates who have completed substantive AI-training work involving earth, climate, environmental, or geospatial content.
What You’ll Work On
• Create realistic computational earth-science tasks that challenge advanced AI models.
• Develop task instructions, reproducible environments, reference solutions, and evaluation criteria.
• Evaluate model-generated code, geospatial analyses, simulations, forecasts, and technical deliverables.
• Identify issues involving coordinate systems, spatial or temporal data, model assumptions, uncertainty, numerical methods, or reproducibility.
Required Qualifications
• 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 relevant earth or environmental-science specialization.
• Hands-on experience using computational methods for geospatial analysis, modeling, simulation, forecasting, remote sensing, statistics, or reproducible research.
Relevant specializations may include geoscience, geology, geophysics, atmosphere and climate, meteorology, ocean and marine science, hydrology, remote sensing, environmental science, and sustainability.
Public Proof of Skill
Strong applicants should provide evidence such as:
• Peer-reviewed publications or preprints involving computational methods
• Publ