Yotta Launches India's First Sovereign B2C AI Chatbot 'myShakti'
Yotta Data Services launched 'myShakti', India's first ever sovereign B2C generative AI chatbot, part of the country's efforts to build up its AI capabilities under its 'Make in India' initiative. Constructed using the open-source AI model of DeepSeek, the 'myShakti' runs fully on Indian servers that guarantee complete data security sovereignty. This development is a major step forward for India in creating indigenous AI models that are not only secure but also affordable.
MyShakti' will thus compete within the recent revised timeline and statements from Union Minister of Railways and Electronics and IT, Ashwini Vaishnaw, who confirmed that India will soon introduce its own safe and secure indigenous AI model within the next six months. While this beta version chatbot is available free for the users on a web app, it will potentially emerge as a significant tool for each user across India.
Developed within four days by a focused team at Yotta, 'myShakti' is hosted on the company's NM1 data center infrastructure. It is running on an advanced setup with 16 nodes of H100 GPUs, accounting for 128 H100s, thus being an efficient and high-performing AI solution. Using Indian servers ensures that all the data processed is within the country's borders.
Darshan Hiranandani, CEO of the Hiranandani Group and Chairman and Co-founder of Yotta, emphasized the importance of 'myShakti' in making India self-reliant in AI. He further stated that hosting DeepSeek in India ensures sovereignty over the AI infrastructure of the country. Co-founder, CEO, and MD of Yotta, Sunil Gupta, reinforced that the company is committed to making cutting-edge AI accessible, secure, and scalable for every Indian.
The minister praised this development, asserting that the model would help indigenous AI to promote India as an ethical technological superpower. Some of the endeavors taken by the country to boost its AI infrastructure are the inauguration of an AI Safety Institute as well as AI model safety deployment with the help of the above.
