AI Analyst (LLM Practice) at Innodata Inc.
Performed human evaluation of AI-generated responses with focus on factual accuracy, relevance, coherence, safety, and policy compliance. Conducted fact-checking and hallucination detection and applied annotation guidelines to support consistent labeling quality. Reviewed model outputs and delivered structured human feedback to improve LLM reliability and training data outcomes. • LLM response safety and policy compliance evaluation • Hallucination detection and ground-truth validation • Content labeling and classification for AI training datasets • QA on model performance metrics and response ranking