India is beginning to take a more grounded approach to artificial intelligence, with a stronger focus on local needs and real-world use.
A fresh collaboration now highlights that direction, especially when it comes to language and accessibility.
The BharatGen Technology Foundation has signed a memorandum of understanding with Indian Institute of Technology Bombay and Centre for Development of Advanced Computing to strengthen foundational AI research. The agreement was formalised on March 17, marking C-DAC’s 39th Foundation Day.
The effort is largely aimed at building AI that can better reflect India’s diversity. Language remains a major hurdle, with many existing systems struggling beyond a handful of widely spoken tongues. This initiative looks to bridge that gap.
BharatGen, which operates out of IIT Bombay with government backing, is developing AI models that can support more than 22 Indian languages. The work goes beyond text, covering speech and document understanding as well, making these systems more practical for everyday applications.
C-DAC contributes its long-standing experience in multilingual computing and high-performance infrastructure, while IIT Bombay provides the research base driving the initiative. Together, they are working toward building a stronger and more self-reliant AI ecosystem.
The partnership will cover several areas, including infrastructure for large-scale AI development, multimodal models that combine text, voice, and visual inputs, and solutions for languages that are often left behind, particularly tribal and low-resource ones. It will also involve creating datasets and evaluation frameworks to support future progress.
A key aspect of this initiative is its focus on inclusion. By bringing lesser-represented languages into the fold early, the effort could help make AI more accessible across different communities.
Alongside this, BharatGen is also building out its computing capabilities. It has partnered with L&T Semiconductor Technologies and L&T Vyoma to develop a complete AI compute platform within India. This includes plans for custom AI chips, a large data centre in Kanchipuram, Tamil Nadu, and a range of AI models tailored to Indian use cases.
The presence of India’s Principal Scientific Adviser at the signing reflects the importance of the initiative. It signals a broader push to ensure that future AI systems built in the country are both locally relevant and capable of scaling globally.
Also Read: BharatGen, L&T Unite to Build Sovereign AI Compute Infrastructure for India
