AI Training Data Evaluator - Project Hedgehog
Currently contributing to Project Hedgehog, an AI training data project focused on evaluating and improving model-generated text outputs. My work involves reviewing submitted responses, comparing alternative outputs, and labeling them based on instruction-following, relevance, clarity, quality, and overall usefulness according to detailed project guidelines. This requires strong judgment, consistency, and close attention to detail to ensure high-quality annotations. The project operates under strict quality standards, handling-time expectations, and workflow guidelines, which has strengthened my ability to perform accurate response evaluation in a structured production environment. I work efficiently while maintaining annotation quality, follow project-specific standards carefully, and help support the creation of reliable training data used to improve AI systems.