AI Data Trainer & Content Evaluator (Remote/Freelance)
Trained and guided AI models by evaluating and ranking model-generated responses for accuracy, helpfulness, and safety compliance. Created high-quality prompts to probe model behavior across domains and documented nuanced preference drivers with analytical justifications. Ensured dataset outputs followed project-specific taxonomies and conversational tone/persona requirements. • Evaluate and rank responses against factual and safety constraints • Design prompt sets for creative and technical reasoning tests • Annotate datasets for supervised fine-tuning with taxonomy adherence • Provide detailed written justifications for model preferences