Generative AI Content Quality Evaluator (Project)
Developed an automated human-in-the-loop evaluation pipeline for assessing LLM-generated content quality. Designed custom metrics for coherence, factuality, and relevance, processing more than 10K outputs daily. Integrated iterative feedback collection to reduce content defect rates and improve model performance.• Designed and executed content evaluation frameworks for LLM outputs • Led human-in-the-loop assessment and quality rating for GenAI data • Established improvement metrics and analyzed review outcomes • Enhanced LLM dataset quality via iterative evaluation cycles