The world’s most popular AI models today are mainly trained using data in the English language, with an an Anglo-centric lens. It is critical to have large language models (LLMs) trained in Indian language data, tailored to understanding Indian nuances, and generating text in Indian languages