Project Hedgehog - AI Response Evaluator
Evaluated AI model outputs across accuracy-focused tasks to improve response quality, factuality, and instruction-following. Reviewed prompts and completions using task-specific rubrics, identified hallucinations, reasoning errors, ambiguity, and low-quality responses, and provided structured feedback to support model improvement. This project strengthened my skills in AI training data review, rubric-based evaluation, attention to detail, written reasoning, and quality assurance for AI systems.