Daniel A.

AI Data Trainer & Researcher

TextRLHF

As an AI Data Trainer & Researcher, I designed and evaluated prompts to refine AI model outputs in complex, nuanced narratives. My work included auditing AI-generated text for factual integrity, logical consistency, and human-centered qualities, such as handling irony and humor. I specialized in comparing AI outputs to primary sources and grading responses for tone, safety, and helpfulness. • Developed and tested multi-layered prompts for investigative storytelling. • Acted as a human filter to assess and improve tone and sentiment understanding in AI models. • Cross-checked model content with original sources to eliminate hallucinations. • Focused on advanced reasoning topics, including law, humor, and blockchain security.