AI Prompt Evaluator
Performed high-fidelity evaluation of text-to-image generative models by assessing the alignment between complex natural language prompts and AI generated visual outputs. I specialized in identifying prompt adherence, visual artifacts, and safety compliance, providing granular feedback to improve model accuracy and reduce hallucinations. My role involved ranking model outputs and performing object detection tasks to ensure structural integrity in generated images. By applying strict quality benchmarks, I contributed to the refinement of datasets used for training large scale computer vision and multimodal AI models.