AI Trainer, Remotasks — Project Dolphin (prompt engineering and AI response evaluation)
Worked as an AI trainer by creating diverse prompts and reviewing AI-generated responses for quality across multiple criteria. Conducted comparative analysis of different model outputs to identify strengths, weaknesses, and actionable improvements. Supported iterative model improvement workflows by structuring evaluation and feedback signals. • Created challenging, varied prompts for AI projects. • Evaluated responses against multiple criteria and metrics. • Compared outputs to highlight gaps and improvement opportunities. • Informed model fine-tuning via structured evaluation notes.