I am the Founder and Executive Lead of AI Safety Nepal (https://www.aisafetynepal.org/), where I work toward building a community of researchers, professionals, and students in Nepal who understand AI safety deeply, contribute to research, and coordinate on governance decisions. I am also currently researching quantifying failure modes in training-time AI alignment techniques. My research interests span AI safety, multilingual NLP, and low-resource language technologies. Previously, I worked as a machine learning engineer at REG-1, building multilingual legal AI systems and domain-specific NLP models for regulatory documents. I graduated in computer engineering from Tribhuvan University, Pulchowk Campus, where I also served as editor-in-chief of The Zerone, the annual technical publication of Pulchowk Campus. My past work includes research on an extractive Nepali question-answering system and building a multilingual legal search platform aimed at improving access to legal information in Nepal.
© Yunika Bajracharya