Bilingual AI Trainer / Data Annotation Specialist
As a Bilingual AI Trainer and Data Annotation Specialist, I worked extensively on enhancing LLM performance via RLHF. My tasks included prompt engineering, adversarial testing, domain benchmarking, project-based data curation, multimodal data labeling, and quality assurance. The projects required nuanced bilingual work in both English and Hindi, often in remote collaboration environments. • Developed and refined prompts in English and Hindi for evaluating LLMs across linguistic and cultural constraints. • Conducted SxS comparisons, stress-testing personalization and memory consistency in models. • Performed adversarial 'Red Teaming' and specialized NEET (Medical Entrance) benchmarking. • Labeled multimodal datasets including image recognition and text categorization, while reviewing responses for alignment and truthfulness.