2 Sources
[1]
Bhashini migrates to Yotta's sovereign AI cloud; sees 40% performance boost, 30% cost dip - The Economic Times
AI-based multilingual translation platform Bhashini has completed its migration from a global hyperscaler to Yotta Data Services' domestic sovereign AI cloud, a move that has enhanced its performance by nearly 40% and reduced operational costs by up to 30%, top officials said. The migration to
[2]
Yotta and BHASHINI Collaborate to Enable Sovereign AI Cloud, Showcasing India's Readiness for Population-Scale AI
* The joint initiative establishes a national reference architecture for Indian organizations transitioning to sovereign, population-scale AI platforms. Yotta Data Services, India's leading sovereign cloud infrastructure and platform services provider, today announced the successful deployment of
Share
Copy Link
India's AI-powered multilingual translation platform Bhashini has completed its migration from a global hyperscaler to Yotta Data Services' domestic sovereign AI cloud. The transition delivered nearly 40% performance improvement and up to 30% cost reduction while ensuring critical AI workloads and data remain within India's jurisdiction.
Bhashini, the AI-based multilingual translation platform developed by MeitY under the National Language Translation Mission, has successfully migrated from a global hyperscaler to Yotta Data Services' domestic sovereign AI cloud infrastructure
1
. The AI cloud transformation marks a significant advancement in India's push for digital sovereignty, ensuring that critical AI workloads and language datasets remain within the country's jurisdiction. The migration was showcased at 'The India AI Sovereignty Dialogues', an official pre-summit event of the AI Impact Summit 20262
.
Source: CXOToday
The transition to Yotta's Government Community Cloud and Shakti Cloud delivered measurable improvements across multiple metrics. Sunil Gupta, cofounder and CEO of Yotta Data Services, confirmed the platform achieved a performance boost of almost 40% while simultaneously reducing operational costs by 20 to 30%
1
. The deployment also maintained 99.99% uptime with zero data loss across more than 200 TiB of data and over 3.5 billion files2
. This wasn't a simple "lift and shift" exercise but required complete architectural transformation, with teams re-engineering multiple layers of the stack.The new architecture implements high-performance GPU clusters powered by NVIDIA H100 processors with optimized pipelines for speech and language models
1
. The platform shifted to an open-source, containerized, and cloud-agnostic architecture, making it completely self-sufficient and vendor-neutral. Amitabh Nag, CEO of Digital India Bhashini Division, revealed the platform has completed approximately 5 billion inferences since its 2024 launch and currently handles about 15 million translation requests daily1
. The system serves diverse use cases ranging from patient-doctor communication to government services.Related Stories
The sovereign AI cloud infrastructure underwent rigorous testing at Maha Kumbh 2025, the world's largest religious gathering, where Bhashini's multilingual AI services operated at population-scale
2
. The platform delivered real-time translation and voice-based assistance in 11+ Indian languages, including the multilingual assistant 'Kumbh Sah'AI'yak'. This deployment demonstrates India's readiness to handle massive AI inferencing workloads on indigenous infrastructure without compromising performance or reliability.The same sovereign cloud architecture hosting Bhashini is now being adopted across the government ecosystem, including by the National Informatics Centre and the Software Technology Parks of India
1
. Gupta cited DOT's International Incoming Spoofed Calls Prevention System running on the cloud, which blocked nearly 1.5 crore spoof calls within the first 24 hours of its October launch. The migration, executed over a two-to-three-month period, covered Bhashini's complete AI stack, including multilingual datasets, models, APIs, containerized services, orchestration pipelines, databases, and storage2
. This creates a modular and reusable reference framework for hyperscaler-to-Indian-cloud transitions across ministries and public sector units, reinforcing India's ambition to position AI as a secure, inclusive public utility supporting economic growth and data sovereignty objectives.Summarized by
Navi
19 May 2025•Technology

28 May 2025•Technology

07 May 2026•Technology

1
Technology

2
Technology

3
Science and Research
