As a Freelance AI Content Moderator & Urdu Annotator, I evaluate Urdu language model responses for truthfulness, helpfulness, and safety.
I flag hate speech, misinformation, self-harm, and culturally sensitive content with South Asian contexts in mind.
I formulate Urdu prompts and golden responses for specialized domain LLMs, then analyze linguistic, factual, and safety failures in model outputs.
In my clinical and technical data work, I verified medical text for factual accuracy and translated complex medical guidelines into accessible Urdu without losing technical precision.
