AI Trainer & Content Evaluator (Freelance)
As an AI Trainer & Content Evaluator (Freelance), I designed and refined multi-turn prompts to enhance large language models’ abilities on various platforms. I evaluated AI-generated outputs based on factual accuracy, coherence, tone, instruction-following, and safety, using strict quality rubrics and performed RLHF-style ranking. I generated high-volume annotated datasets, documented evaluations, and applied red-teaming methods to detect harmful or inaccurate responses. • Crafted and refined prompts to stress-test models and explore edge cases • Used ranking and annotation to provide supervision signals for model learning • Evaluated outputs using established rubrics for quality, safety, and factual correctness • Maintained structured documentation and reported failure modes in model responses