AI Training Data Evaluator & Annotation Workflow Designer
Contributed to multiple AI training data projects across freelance platforms and independent workflows, focusing on evaluation, annotation, and refinement of large language model (LLM) outputs. Work included classification, segmentation, and ranking tasks, as well as qualitative assessment of generated responses based on accuracy, coherence, tone, and contextual appropriateness. I've also designed and iterated on custom annotation workflows and prompt structures to improve output quality and consistency. Identified gaps in task instructions, flagged edge cases, and contributed to refining evaluation rubrics to better capture nuanced or subjective criteria. Regularly balanced speed with precision while maintaining high inter-rater alignment, particularly in projects involving complex language, sensitive topics, or human-centered content.