Odisha, Tamil Nadu Join Hands with Sarvam AI for Population-Scale AI Deployment

Sarvam AI recently announced strategic partnerships with Odisha and Tamil Nadu governemnts to build sovereign AI infrastructure and drive population-scale adoption across public services.

The partnerships aim to establish at-scale compute facilities, develop sovereign models for Indian languages, and build institutional capacity for AI deployment in governance.

The announcement from the Bengaluru-based AI infrastructure company marks a significant development in India’s sovereign AI journey, with both state governments committing to building AI-optimized datacenters and deploying language-first applications across citizen services.

Odisha: 50MW AI Capacity Hub

The Odisha government signed an MOU with Sarvam to build a sovereign AI capacity hub featuring a 50MW AI-optimized facility. The infrastructure is designed to function as the state’s own AI public utility while also positioning Odisha as India’s compute backbone for other states and national platforms.

The facility will initially focus on the state’s core strengths in mining, heavy industrials, and skilling. Vision AI systems will be deployed to ensure safety and compliance at industrial sites, while Odia-to-English voice tools are planned to scale job readiness for youth across the state.

"Odisha’s transition from a mine-driven economy to a mind-driven economy, a technology-enabled economy, is aligned with the broader vision of Viksit Bharat 2047," said Smt. Anu Garg, Chief Secretary of Odisha.

Beyond serving Odisha, the hub is positioned to anchor a nationwide compute grid. Sarvam stated that the facility will offer reliable, production-grade AI capacity to address India’s compute constraints as the country scales AI adoption.

Tamil Nadu: Digital Sangam Research Park

Tamil Nadu is establishing Digital Sangam which is described as India’s first Sovereign AI Research Park, through a public-private partnership involving Sarvam AI and IIT Madras. The park will feature a 20MW AI-optimized datacenter and serve as a dedicated ecosystem where advanced compute infrastructure, frontier research, and startup incubation co-locate.

The sovereign infrastructure is designed to ensure sensitive data remains within national boundaries while enabling AI systems that reflect Tamil language and culture. Sarvam AI brings experience in building population-scale AI systems, while IIT Madras anchors research and talent development, it said in a release.

"The deployment of population-scale, voice-enabled AI applications will enable citizens, particularly in rural and tribal areas, to access information on entitlements, benefits, and grievance redressal through simple conversational interfaces," said Dr. TRB Rajaa, Minister for Industries, Tamil Nadu.

The partnership plans to deploy AI across citizen services, including a Vivasaya Nanban assistant to provide 24x7 digital advisory to 79 lakh farm households and a unified AI-powered citizen helpline to simplify access to welfare and public services.

Sovereign AI Infrastructure Vision

According to Sarvam, the partnerships represent the first components of a national compute grid designed to support India’s workforce at scale. The company emphasized that building sovereign AI capacity is central to a digitally self-reliant economy, requiring deliberate architecture including sovereign compute, models built for Indian languages and contexts, and a Digital Public Infrastructure layer that allows intelligence to be shared and reused across systems.

Sarvam stated it is building India’s full-stack sovereign AI infrastructure end-to-end, from models grounded in Indian data to applications deployed across population-scale use cases. It said that the goal is to ensure that AI systems reflect how India thinks, speaks, and solves complex problems, while ensuring learning, value, and capability remain within national boundaries.

The announcements come as India seeks to establish digital autonomy in AI infrastructure. The partnerships position state governments as active participants in building sovereign compute capacity, with implications for how AI infrastructure scales across the country in the coming years.