Chemistry / Chemical Engineering AI Evaluation Specialist — Remote AI Evaluation Projects
Remote contractor providing expert review for AI training problems in Chemistry and Chemical Engineering. Reviewed prompts, reference answers, grading rubrics, and research-style task environments to assess scientific realism, internal consistency, correctness, and grading fairness. Produced detailed annotations and reviewer feedback to improve dataset quality, rubric reliability, and prompt clarity. • Evaluated LLM responses for chemistry accuracy, reasoning quality, completeness, and alignment with accepted principles • Validated reference answers and whether scoring logic rewards correct reasoning and penalizes incorrect conclusions • Identified scientific issues such as hallucinated mechanisms, unsupported claims, incorrect equations, and unrealistic laboratory conditions • Completed expert review tasks typically taking 30–90 minutes while following strict guidelines and QA standards