Remote AI training and data labeling jobs

USPTO registered patent agents write and grade claim drafting and prosecution scenarios used to test AI models. The board headlines $108/hr; the ad states $90 to $125/hr, weekly via Stripe, remote.

Licensed PEs write design and review scenarios for AI evaluations, with code and safety reasoning, and grade model output for code compliance. $105 to $190/hour (board headline $148), paid weekly via Stripe, remote, minimum 10 hours a week.

Senior hardware experts (CAD, RTL, robotics; 8–12 years) design problem sets, reference solutions and rubrics to evaluate frontier AI models. Board shows $135/hr; the ad gives $100–170/hr by experience. Remote, async research work, not production engineering.

ML researchers with computational materials science expertise build and evaluate models on simulation data (DFT, molecular dynamics, Monte Carlo). Remote project work over 2 to 3 weeks, 10 to 20 hours a week, $150 to $200/hour (board headline $175/hour).