AI Model Evaluator & RLHF Trainer
As an AI Model Evaluator & RLHF Trainer at Deccan AI, I participated in reinforcement learning from human feedback (RLHF) pipelines to enhance AI model quality. My daily tasks included evaluating model outputs for performance, accuracy, and reasoning, and delivering comprehensive feedback. I consistently identified factual errors, inconsistencies, and poor reasoning within AI-generated responses. • Compared multiple model-generated responses and selected the best output based on defined criteria. • Collaborated with the AI development team to refine prompt guidelines and evaluation rubrics. • Contributed to data annotation processes that supported continuous AI improvement. • Ensured high labeling quality by adhering to standardized evaluation protocols.