Data Annotation Specialist
I worked as a Data Annotation Specialist on AI training projects, focusing on image annotation and prompt-response writing (SFT) to enhance model performance, accuracy, and safety. My role involved labeling and categorizing data based on defined taxonomies, as well as creating high-quality prompt-response pairs for fine-tuning AI models. In addition to annotation, I actively contributed to AI security by identifying harmful, biased, or unsafe outputs and ensuring compliance with safety guidelines. I performed quality assurance reviews, detected inconsistencies, and applied validation techniques to improve data reliability and reduce errors. With my background in cybersecurity and penetration testing, I brought a unique perspective to securing AI systems against vulnerabilities, making the training data both accurate and robust for real-world deployment.