Hamza Humayun

I am an AI Trainer, QA Annotator, and Python developer with deep experience in AI model evaluation, data annotation, and rubric-driven quality assurance. I design and execute test cases, perform regression testing, and document defects to improve model robustness and reliability.

I have built backend automation and API integrations for healthcare and government projects, delivered 15+ Python solutions, and developed computer vision systems such as face detection, ANPR, and gesture recognition. I consistently maintain high annotation accuracy (95%+–98%+) and inter-rater reliability (Cohen's Kappa > 0.85).

I collaborate with cross-functional teams to refine evaluation frameworks, optimize annotation workflows, and reduce annotation time by 30% while preserving quality. I bring a rigorous, analytical approach to advancing AI capabilities through testing, annotation, and continuous improvement.