AI Evaluation/Scenario Writing
Contributed to AI training initiatives involving both image annotation and text-based evaluation tasks. For computer vision datasets, I performed bounding box annotation on images containing multiple objects, ensuring tight box placement, accurate class labeling, and compliance with YOLO formatting requirements. I followed structured guidelines to maintain consistency and conducted quality checks prior to dataset export. In addition, I supported natural language processing projects by labeling and classifying text data for sentiment and intent analysis. I evaluated AI-generated responses using structured rubrics, assessing tone, relevance, logical consistency, and policy adherence. I identified inconsistencies, flagged ambiguous cases, and ensured alignment with defined quality standards to support reliable model training and performance improvement.