Google CRC Project( Bangla Language ) ,_preferences_multilingual and more
Worked on multiple large-scale AI data annotation and evaluation projects for leading platforms, contributing to the development, safety, and multilingual performance of advanced language models. Projects included: Google CRC Languages Preference Ranking and Rewrites Red Teaming (M++) i18n Canonical Safety Multilingual Preference Evaluation And other AI training and evaluation initiatives Key Responsibilities & Labeling Types: Performed data segmentation and content categorization across diverse datasets Conducted evaluation and rating of model outputs based on quality, relevance, safety, and linguistic accuracy Worked on Prompt + Response Writing (Supervised Fine-Tuning – SFT) to improve model alignment and response quality Ranked and rewrote multilingual responses to reflect human preferences and cultural nuances Participated in red teaming tasks to identify safety risks, edge cases, and failure modes in AI systems