Yotta Data Services has launched myShakti, the country’s first fully sovereign B2C generative AI chatbot that runs off DeepSeek Open-Source AI model. Developed by Yotta’s team within 4 days, the chatbot is entirely hosted in India.

The chatbot is deployed within NM1, Yotta’s Mumbai data centre, the only multi-tenant Tier IV data centre facility, on an advanced server infrastructure comprising 16 nodes of Nvidia’s H100 GPUs—a total of 128 H100s. The product is currently available as a beta version on a web app and is free to use.

“myShakti brings DeepSeek’s Open-Source AI model completely to Indian servers (and disconnected from any other country) addressing the nation’s data security and privacy concerns. myShakti shall enable all open-sourced and partnered data sets to be processed on servers located within India only,” the company said in a press release.

A sovereign B2C gen AI chatbot is a generative AI-powered chatbot designed for business-to-consumer (B2C) interactions while operating under sovereign cloud or data regulations.

The company has also integrated a comprehensive feedback mechanism to refine and improve myShakti capabilities over time.

“By lowering barriers to entry, we are democratizing AI, enabling startups, businesses, and researchers to innovate without the constraints of high infrastructure costs,” said Sunil Gupta, Co-founder, CEO & MD, Yotta Data Services in a statement.

“This could be India’s Sputnik moment for AI—an inflection point that propels the nation to the forefront of global AI innovation. As adoption scales, platforms like Yotta’s Shakti Cloud will provide the critical infrastructure to power this transformation,” he added.

Yotta operates cloud regions at its hyperscale data centre parks in Panvel (Navi Mumbai) and Greater Noida (Delhi NCR). Its sovereign hyperscale cloud, Yntraa, is open-source-based, MeitY-empanelled (VCC and GCC), and deployed in government-owned CSPs under a white-label/PPP model. Yotta has also launched Shakti Cloud, offering AI services such as AI labs, workspaces, serverless inferencing, Nvidia's NIM services, and Kubernetes clusters with GPU resources.