What if your expertise in computational fluid dynamics could teach AI to reason about aerodynamics, boundary conditions, and flow physics — without you ever having to write a single line of machine-learning code?
We're looking for CFD engineers and aerodynamicists to design small, well-defined engineering challenges that train and evaluate AI models on real fluid-dynamics reasoning. Each task you create has three parts: a transparent problem statement, a deterministic checker that scores the model's answer, and a trusted reference solution that proves the task is solvable. Think of it as writing bite-sized CFD exam problems — grounded in physics, unambiguous, and automatically gradable.
No reinforcement-learning or AI research background is required. What matters is your hands-on simulation experience and your ability to translate domain knowledge into precise, solvable tasks.
self-contained CFD and aerodynamics tasks that test an AI model's engineering reasoning
Write clear, unambiguous problem statements covering topics such as boundary-condition selection, mesh quality assessment, flow-regime identification, and simulation setup
Develop deterministic scoring checkers that can automatically evaluate a model's final answer
Provide verified reference solutions that confirm each task is solvable and correctly scored
Review and iterate on tasks to ensure physical accuracy, clarity, and appropriate difficulty
Draw on your experience with OpenFOAM to ground tasks in real-world practice
Work independently and asynchronously — fully on your own schedule
Who You Are
Hands-on experience with OpenFOAM
Solid foundation in fluid mechanics, aerodynamics, and heat transfer fundamentals
Comfortable with mesh generation, turbulence modeling, boundary-condition specification, and post-processing
Able to formulate