Mechanical Engineer & Python Expert – Freelance AI Trainer

About the role

You will design computational engineering problems that challenge frontier AI models. Each problem must be solvable with a specialized tool (e.g., Cantera, CoolProp, CalculiX, OpenFAST) and have a verifiable answer generated by code. Work is project‑based, runs inside sealed Linux containers, and is compensated per task.

Key responsibilities

• Select an anchor simulation tool and create a problem that relies on its solvers or domain‑specific models.
• Write a reference Python solution, provide necessary input files, geometry, or mechanism definitions.
• Define the numerical answer and tolerance required for the AI model to be considered correct.
• Test the problem against the AI model in parallel batches, adjusting difficulty to achieve a 10‑30% success rate.
• Iterate on thermodynamic cycles, material models, and boundary conditions to fine‑tune calibration.
• Submit the finalized task to a senior reviewer for quality assurance.

Required profile

• Degree in Mechanical Engineering or a related field.
• At least 2 years of research, applied, or teaching experience.
• Strong Python programming skills for writing reference solutions.
• Fluency with, or willingness to quickly learn, at least one scriptable mechanical‑engineering package (e.g., Cantera, CoolProp, CalculiX, OpenFAST, YADE).
• Ability to design problems that genuinely require a specialized solver.
• Excellent written English (C1 level or higher).

Required skills

• Python
• Cantera, CoolProp, CalculiX, OpenFAST, or similar engineering simulation tools
• Linux container environments
• Computational problem design
• Thermodynamics, fluid mechanics, structural analysis basics

What we offer

• Project‑based freelance contracts with payment per completed task.
• Flexible, remote work schedule.
• Opportunity to deepen expertise in advanced simulation tools and AI model evaluation.