RT team
Former Tech Mahindra CEO, CP Gurnani, has achieved a significant milestone by developing an Indian large-language model (LLM) for under $5 million. This accomplishment, known as “Project Indus,” was shared by Gurnani at the fifth annual MachineCon GCC Summit in Bangalore. The project, completed in just five months, enables communication in about 40 local languages and dialects.
The initiative was driven by a challenge from Sam Altman of OpenAI, who doubted India’s ability to compete in the LLM space with a limited budget. At an event organized by the Economic Times, Altman implied that India couldn’t train foundation models with only $10 million. This skepticism spurred Gurnani to act swiftly, and within six hours of discussing with his chief innovation officer, a plan was set into motion.
“I am pleased to announce that we achieved what Sam Altman thought impossible, spending less than $5 million,” Gurnani stated. Although Altman later clarified his comments on social media, encouraging innovation and unique contributions rather than direct competition, Gurnani’s response highlights India’s growing prowess in AI development.
Project Indus is designed to revolutionize AI communication in India by supporting numerous Indic languages and dialects. Nikhil Malhotra, Tech Mahindra’s global head of Makers Lab, detailed the extensive data collection efforts involved, particularly from Hindi-speaking populations. The project utilizes Dell Technologies’ “GenAI in a box” framework, ensuring easy enterprise deployment and scalability. Denise Millard of Dell emphasized the need for accessibility in AI solutions, which this collaboration aims to provide across sectors like healthcare, rural education, banking, finance, agriculture, and telecom.
Additionally, Tech Mahindra plans to expand Project Indus to preserve Bahasa Indonesia, allowing communication in local dialects online. This extension demonstrates the model’s versatility and broader impact.
Gurnani’s announcement marks a crucial moment for India in the AI arena, showcasing the nation’s capability to develop and implement advanced AI technologies effectively.
