Loading...
Mercor is building the Industrial Technical Visual Reasoning Benchmark, an evaluation dataset that tests how well AI models reason over authentic engineering visuals. We are hiring Utilities & Power Systems experts to author benchmark problems in your field.
For each problem you will source a genuine professional artifact (grid diagrams, electrical diagrams, substation layouts), write a problem that cannot be solved from the text alone, determine a concise and unambiguous final answer, and write a step-by-step worked solution at a professional level. You will tag each problem with its domain, artifact type, skill tested, and difficulty. Every problem is independently reviewed by another domain expert before it ships.
This is a part-time, remote, hourly engagement for practicing professionals with hands-on experience reading real power systems and utilities artifacts. Simplified textbook figures and questions answerable by simple lookup are out of scope; the work rewards genuine structural, spatial, and multi-step reasoning.
Apply directly on Mercor to get started.