Next month, a group supported by India’s top engineering colleges and Reliance Industries Ltd., owned by Mukesh Ambani, plans to launch its first ChatGPT-style service. This is a significant step toward the nation’s goal of becoming a major player in the artificial intelligence space.

During a technology conference in Mumbai on Tuesday, the BharatGPT group—which includes eight affiliated colleges and an arm of the most valued firm in India—offered a preview of the huge language model. A developer in Hyderabad utilised the technology to write computer code, a banker spoke with it in Hindi, and a motorbike repairman in southern India asked an AI bot questions in his native Tamil in a video that was shown to the delegates.

If the model is effective, it will be a step forward for India in the rapidly progressing race to build potentially revolutionary AI technology. Hanooman is named after the Hindu deity who is half-monkey. According to BharatGPT, the model will operate in four primary domains—health care, governance, financial services, and education—using eleven indigenous languages. The approach was created in partnership with Indian Institute of Technology universities, notably those in Bombay, with support from the Indian government and wireless carrier Reliance Jio Infocomm Ltd.

With support from well-known VC firms like Lightspeed Venture Partners and billionaire Vinod Khosla’s fund, a number of businesses, like Sarvam and Krutrim, are also developing open-source AI models specifically tailored for the Indian market. Even though OpenAI and other Silicon Valley corporations are creating ever-larger LLMs, these efforts are hampered by computing limitations and the need for simpler models that are more inexpensive for smaller enterprises and government agencies.

According to Ganesh Ramakrishnan, chair of the computer science and engineering department at IIT Bombay, “it’s a different genre of LLMs.”

In an interview conducted on the fringes of the yearly Nasscom IT industry conference, Hanooman stated that it will also provide speech-to-text functionality, making it much more user-friendly. Millions of people in this 1.4 billion-person nation are illiterate.

He stated that Reliance Jio will construct unique models for particular applications. The telecom-to-retail giant is already developing Jio Brain, a platform that will apply artificial intelligence to a network of over 450 million users.

Large-scale machine learning (LLM) systems produce responses that seem natural by learning from large amounts of data. These models make use of generative AI, a more recent branch of artificial intelligence made well-known by OpenAI’s ChatGPT.

Because it’s the first private-public partnership of its sort in the nation and incorporates significant participants in various industries, BharatGPT’s endeavour in this area is a little different from others.

In reference to the multigenerational family systems that are still prevalent in India, Ramakrishnan remarked, “It’s like the Indian joint family.” “We function better together because we are interdependent.”