AI Trainer (Freelance) | Outlier AI
As an AI Trainer at Outlier AI, I design, evaluate, and rate large language model conversational prompts, focusing on reasoning, factual accuracy, and instruction-following. I perform structured side-by-side response evaluation and write clear rationales to document model behaviors and identify areas for improvement. My work leverages bilingual Spanish–English expertise to assess dialect accuracy, idiomatic correctness, and cultural relevance of AI outputs. • Rate and rank AI-generated responses and check for grounding errors. • Execute prompt engineering and fact-checking workflows. • Critically review AI model outputs for hallucinations and tone mismatches. • Apply deep knowledge of Latin American Spanish for language-specific LLM evaluation.