Materials Engineer & Python Expert - Freelance AI Trainer
You design computational material science problems to challenge a frontier AI model with problems that have verifiable answers by code and require specialized tools like ObsPy, instaseis, pyrocko, MITgcm, flopy/MODFLOW, or others. You pick an anchor tool and design a problem focusing on its waveform-processing kernels, geophysical inversion routines, sub-surface flow solvers, or community-validated data pipelines. You write a Python reference solution, supply input files and model or domain definitions where necessary, decide the numerical answer and required tolerance, test the problem against the AI model in batches of parallel attempts, tune the problem difficulty to achieve a low pass rate, and submit the task to a senior reviewer for quality feedback. Calibration involves tuning the problem by rewriting scenarios, tightening parameters and solver tolerances, and observing the model's behavior, which builds deeper command of the tool and understanding of how the AI model navigates complex geophysical problems.
Mechanical Engineer & Python Expert - Freelance AI Trainer
Design computational engineering problems to challenge a frontier AI model using specialized tools like Cantera, CoolProp, CalculiX, OpenFAST, or others installed in a sealed Linux container. Write Python reference solutions and supply necessary input files and definitions. Determine the numerical answer and appropriate domain-specific tolerance for correctness. Test and tune the problem difficulty against batches of parallel model attempts to achieve a pass rate between 10-30%. Submit tasks for review by a senior expert for feedback and quality assurance. Continuously refine problems by rewriting thermodynamic cycles, adjusting material models and boundary conditions, and analyzing model behavior through test attempts. Gain a deeper understanding of both the engineering tools and the AI model's approach to complex thermal, structural, and fluid mechanics problems.
Mechanical Engineer & Python Expert - Freelance AI Trainer
Design computational engineering problems to challenge a frontier AI model, ensuring the problem has an answer verifiable by code and requires a specialized tool like Cantera, CoolProp, CalculiX, OpenFAST, or others. Each problem runs inside a sealed Linux container with the pre-installed tool and a programmatic judge that grades the model's answer. Tasks include picking an anchor tool and designing a problem that hinges on its solvers, simulation kernels, or domain-specific models, writing a Python reference solution, supplying input files and geometry or mechanism definitions as needed, deciding the numerical answer and tolerance for correctness, testing the problem against the model in batches of parallel attempts, tuning the problem difficulty until the agent only succeeds in a small number of attempts, and submitting tasks to a senior reviewer for feedback and quality assurance. Calibration involves tuning the problem against batches of parallel runs to aim for a pass rate in the 10–30% band, rewriting thermodynamic cycles, tightening material models and boundary conditions, and observing how the agents behave.
Mechanical Engineer & Python Expert - Freelance AI Trainer
Design computational engineering problems that challenge a frontier AI model using specialized tools like Cantera, CoolProp, CalculiX, OpenFAST, or similar. Create problems that have verifiable answers through code and require these specialized solvers, simulation kernels, or domain-specific models. Write Python reference solutions, supply necessary input files and definitions, determine numerical answers with domain-appropriate tolerance for correctness, and test problems against the AI model in batches of parallel attempts. Tune problem difficulty so the model succeeds only in a small number of attempts, and then submit tasks to a senior reviewer for feedback. Calibrate the problem by rewriting thermodynamic cycles, adjusting material models and boundary conditions, and observing how the agents behave, aiming for a pass rate between 10–30%. Gain deeper command of the anchor tool and develop intuition for how the AI navigates complex thermal, structural, and fluid mechanics problems.
Mechanical Engineer & Python Expert - Freelance AI Trainer
Design computational engineering problems to challenge a frontier AI model, ensuring each problem has an answer verifiable by code and requires a specialized tool like Cantera, CoolProp, CalculiX, OpenFAST, or others. Each problem runs inside a sealed Linux container with the tool pre-installed and a programmatic judge that grades the model's answer. Pick an anchor tool and design a problem that hinges on its solvers, simulation kernels, or domain-specific models. Write a Python reference solution, supply input files and geometry or mechanism definitions where needed. Decide the numerical answer and determine the acceptable tolerance for the model's correct response. Test the problem against the model in batches of parallel attempts, tuning the problem difficulty until the agent succeeds only in a small number of attempts. After finalizing the task scoring, submit it to a senior reviewer in the subfield for feedback to ensure task quality is high. Tune problems through rewriting thermodynamic cycles, tightening material models and boundary conditions, and monitoring agent performance to maintain a pass rate in the 10–30% range while deepening command of the anchor tool and developing understanding of how frontier models navigate complex thermal, structural, and fluid mechanics problems.
US Corporate Attorney - Freelance AI Trainer
Contributors may generate prompts that challenge AI; evaluate AI-generated solutions for correctness, assumptions, and logic; improve AI reasoning to align with first principles and accepted standards; and apply structured scoring criteria to assess multi-step problem solving.
US Corporate Attorney - Freelance AI Trainer
Contributors may generate prompts that challenge AI, evaluate AI-generated solutions for correctness, assumptions, and logic, improve AI reasoning to align with first principles and accepted standards, and apply structured scoring criteria to assess multi-step problem solving.
US Corporate Attorney - Freelance AI Trainer
Contributors may generate prompts that challenge AI, evaluate AI-generated solutions for correctness, assumptions, and logic, improve AI reasoning to align with first principles and accepted standards, and apply structured scoring criteria to assess multi-step problem solving.
Optical Engineer - Freelance AI Trainer
Design original optics problems that simulate real physics research workflows; ensure problems are computationally intensive and cannot be solved manually within reasonable timeframes (days/weeks); develop problems requiring non-trivial reasoning chains in mechanics, electromagnetism, thermodynamics, and quantum mechanics; base problems on real research challenges or practical applications from optics and physics practice; document problem statements clearly and provide verified correct answers.
Freelance Mathematics Expert - AI Trainer
Design rigorous mathematics problems reflecting professional practice; evaluate AI solutions for correctness, assumptions, and constraints; improve AI reasoning to align with industry-standard logic; apply structured scoring criteria to multi-step problems.
Access all 4,256 remote & onsite AI jobs.
Frequently Asked Questions
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.