HackTalk: Janhavi Khindkar

​HackTalk: Janhavi Khindkar

​Day 2 talk for the Global South AI Safety Hackathon, June 19-21 online. Free and open.

​Janhavi Khindkar is an Applied AI Researcher and Engineer working on Bhashini, India's national multilingual AI platform under MeitY, where she works on model optimization, fine-tuning, and deployment for low-resource Indic languages at scale. She also leads ValueShift Research, an independent AI safety collaboration focused on mechanistic interpretability and AI control. Her work sits at the intersection of applied ML infrastructure and AI safety, with a particular interest in how safety alignment behaves across languages and cultural contexts.

​Format: ~20-30 min + Q&A. Online (Zoom), livestreamed and recorded.

​Part of the Global South AI Safety Hackathon, June 19-21, online with in-person hubs across Latin America, Africa, and Asia. Organized by Apart Research.