AI Content Trainer & Evaluator (Mindrift, Appen)
Served as an AI Content Trainer & Evaluator conducting prompt evaluation, hallucination detection, and response quality assessment. Worked with platforms such as Mindrift and Appen to create and rate large language model (LLM) outputs in English and French. Focused on reviewing content quality, fact-checking, and preference ranking to train AI systems for higher accuracy, clarity, and impact. • Performed regular prompt evaluation and rating based on clarity, tone, and factual accuracy. • Detected hallucinations and misinformation through structured fact-checking workflows. • Conducted response comparison and preference ranking across multiple LLM completions. • Generated and evaluated prompts and structured narratives for training datasets.