Overview
We’re looking for an experienced Mechanical Design Engineer to help train AI models. You’ll write hard, verifiable engineering problems drawn from your own specialty, work out the answers yourself, and show us exactly where today’s AI systems still fall short of a practicing engineer.
This is a new and fast-growing area for our company, focused on teaching AI to reason like an engineer. The models can already talk fluently about mechanical engineering. What they can’t yet do reliably is the actual work: size a component against a real load case, catch the failure mode a spec sheet doesn’t mention, read a CAD assembly and tell you what interferes at full travel, or predict what a design change breaks downstream. Your engineering judgment becomes the benchmark those models are measured against.
What you’ll actually do
- Write challenging, closed-ended problems from your area of expertise, built so that solving them requires real engineering reasoning rather than a lookup.
- Derive the correct answer, verify it, and write a worked solution showing how you got there.
- Test your problems against AI systems and sharpen the ones they get right until they challenge the model.
- Review AI-generated engineering analysis, designs, and technical documentation for correctness and real-world applicability.
- Review other contributors’ problems and solutions for correctness, ambiguity, and rigor.
Problems draw on whatever you know best
- Mechanical design and CAD: geometry, clearances, tolerances, mechanisms, and the effect of design changes.
- Structures and materials: stress, deflection, fatigue, buckling, material selection, and failure analysis.
- Thermal and fluids: heat transfer, thermodynamics, HVAC, pumps and piping, and fluid mechanics.