Materials Scientist for AI Training

Role Overview

Apply materials science and Python expertise to create, solve, review, and validate technical tasks used to improve AI systems. You will assess material structures, properties, processing, performance, and failure through reproducible computational workflows, ensuring results are both computationally valid and physically meaningful.

Key Responsibilities

• Solve and validate computational materials-science and materials-engineering problems.
• Create material structures, atomic configurations, compositions, and solver-ready inputs.
• Model relationships among composition, structure, processing, properties, and performance.
• Run atomistic, electronic-structure, molecular-dynamics, continuum, electrochemical, or related simulations.
• Use Python to generate inputs, automate calculations, run parameter sweeps, process results, and validate outputs.
• Analyze mechanical, thermal, electrical, chemical, structural, and electrochemical properties.
• Diagnose failed calculations, invalid structures, convergence issues, numerical instability, and incorrect physical assumptions.
• Compare computational outputs with experimental data, literature values, known properties, or expected physical trends.
• Review AI-generated solutions for scientific accuracy and identify invalid assumptions, configurations, or conclusions.
• Develop reproducible reference solutions and objective verification methods.

Qualifications

• MS or PhD in Materials Science and Engineering, Metallurgy, or a closely related field, or an MS or PhD in Mechanical Engineering or Chemical Engineering with substantial materials specialization.
• Strong understanding of materials behavior and structure-property relationships.
• Experience in computational materials modeling, simulation, characterization, or materials-focused engineering analysis.
• Practical proficiency with Python.
• Experience with at least one engineering or scientific tool operated through a command-line interface, scripting interface, configuration files, or programmatic API.
• Ability to justify modeling assumptions, parameters, approximations, and convergence criteria, and to distinguish computational failures from genuine physical behavior.
• Ability to clearly explain complex scientific reasoning and technical limitations.
• Experience limited exclusively to graphical user interfaces is not sufficient, because solutions must be reproducible through code, scripts, configuration files, or command-line tools.
• Relevant tools may include LAMMPS, ASE, pymatgen, Quantum ESPRESSO, and FEniCSx.

Work Terms

• Independent contractor engagement, fully remote and open globally.
• Approximately 15 hours per week.
• Flexible schedule, with choice over work hours and days, including weekends if preferred.

Compensation

$80 to $130 per hour.