This role involves training and evaluating AI agents on expert-level computational fluid dynamics tasks. You will set up and run simulations using industry-standard CFD tools, create verified reference solutions, and develop Python validation checks to assess agent performance on fluid simulation problems.
Pay not disclosed
The platform does not publish the pay for this job.
Open to residents of: United States.
You apply on Mindrift, the AI training platform run by Toloka: submit your application for the role, pass a qualification step, then work on projects that match your expertise.
Official posting: https://apply.workable.com/j/36F55997E3
Last checked on October 8, 2026.
This evaluation role involves designing simulations and technical specifications to test the ability of advanced AI models to reason from first principles in your area of expertise in engineering. You will define complex design problems (controllers, circuits, geometries) that the AI must solve by demonstrating reasoning capability rather than simple data retrieval.
Posted August 31, 2026
This role involves designing small, self-contained computational fluid dynamics tasks that train and evaluate AI models on engineering reasoning. CFD engineers create problem statements, scoring mechanisms, and reference solutions to teach AI systems about aerodynamics, boundary conditions, and flow physics, without requiring machine learning expertise.
This role involves creating rigorous mechanical engineering problems and solutions to train AI models to reason like practicing engineers. You will write verifiable problems from your specialty, derive correct answers, test them against AI systems, and review AI-generated engineering work for real-world applicability.
A mechanical designer writes rigorous engineering problems and solutions to train AI models to reason like practicing engineers. You'll create challenging, closed-ended technical problems from your specialty, verify correct answers, test them against AI systems, and review AI-generated engineering work for real-world applicability.
This role involves training and evaluating AI agents on fluid simulation tasks by setting up CFD simulations, creating reference solutions, and validating agent outputs. It is designed for experienced mechanical or aerospace engineers with strong computational fluid dynamics expertise.
Posted October 6, 2026
This remote contract role asks you to build and refine MuJoCo simulation environments that teach AI agents locomotion, manipulation, and multi-agent coordination. It is aimed at experienced robotics and machine learning practitioners who are comfortable with physics simulation, reinforcement learning, and robot control.
