At Innodata India Pvt. Ltd., I annotate and label multimodal datasets to train and improve Large Language Models. I also evaluate AI-generated responses for accuracy, relevance, coherence, and safety.
I rate model outputs against defined quality guidelines and analyze prompts and responses to identify model strengths, weaknesses, and failure patterns. My quality assurance reviews help maintain annotation consistency and dataset accuracy.
Before Innodata, I worked in customer care at Kochartech and Tech Mahindra. I also co-authored “HUMAN-IN-THE-LOOP EVALUATION FOR ERROR AND BIAS REDUCTION IN AI SYSTEMS,” published in the International Scientific Journal of Engineering and Management.
