Freelance AI Trainer & Prompt Engineer (Remote)
Provided prompt and response evaluation for LLM outputs across technical, creative, and domain-specific categories. Ranked model responses using RLHF principles based on truthfulness, safety, and adherence to instructions. Performed fact-checking and hallucination detection to support the creation of high-quality training datasets. • Prompt engineering for LLM performance evaluation • RLHF-based response ranking • Hallucination detection and fact-checking • Urban/planning domain checks for technical and spatial accuracy