The work
You will support an AI training project by solving challenging scientific, mathematical, or technical problems in an area you know well. You will also use Python for calculations or simulations and review AI-generated solutions for accuracy.
Your reviews will look for incorrect assumptions, reasoning errors, missing constraints, calculation problems, and other technical mistakes. You will explain errors clearly and decide whether answers meet the relevant technical standards.
- Solve complex problems in your STEM or scientific specialty.
- Develop rigorous solutions and use Python for calculations or simulations.
- Evaluate AI-generated answers for factual, mathematical, and scientific correctness.
- Identify faulty assumptions, missing constraints, reasoning errors, and subtle technical issues.
- Use sound judgment to assess whether technical answers meet the required standards.
What it pays and takes
This is a fully remote, part-time contractor role with a flexible schedule. The expected workload is approximately 15 hours per week, including the option to work weekends.
- $60 to $120 per hour equivalent, paid per task that meets the project specifications.
- Approximately 15 hours per week on a flexible schedule.
- Remote work open globally in eligible countries, subject to applicable hiring eligibility.
- Professional fluency in English.
- Deep expertise in at least one STEM or scientific field, such as chemistry, biology, mathematics, physics, earth science, geoscience, environmental science, robotics, or a related quantitative discipline.
- Practical proficiency with Python or another programming language.
- Strong scientific, mathematical, or quantitative problem-solving skills.
- Experience using coding agents in a technical workflow.
- Ability to evaluate technical answers precisely and explain errors in reasoning, calculations, or assumptions.
- Experience may come from research, industry, teaching, engineering, or independent technical work. No specific degree or number of years of experience is required when strong expertise can be demonstrated.
- Knowledge of NumPy, SciPy, pandas, SymPy, Jupyter, or other scientific libraries can help, but no particular library is required.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training work
AI training is the human work behind systems that learn from examples, including solving problems and reviewing model answers. People with strong technical expertise help identify accurate reasoning and errors so AI systems can produce more reliable results.