Dilayehu Tamrat

I am an AI data annotation and training specialist with over five years of experience evaluating and training multimodal generative AI systems across text, image, and video. I apply rubric-based evaluation, QA auditing, and model refinement focused on truthfulness, safety, instruction alignment, helpfulness, realism, and consistency.

As a licensed medical doctor and pediatrics resident, I bring active clinical experience and diagnostic reasoning to high-stakes AI validation, fact-checking medical outputs, and translating AI concepts into clinically relevant applications. I have designed and delivered a lecture series on AI in medicine and participate in public health research and evidence-based medical content creation.

I consistently support diverse AI workflows—agent benchmarking, content detection, relevance labeling, moderation, chain-of-thought reasoning, and multimodal visual evaluation—and am frequently selected for QA, auditing, and guideline-improvement roles to enhance inter-annotator agreement and model safety.