BharatGen unveils PARAM 2, a 17 billion-parameter mixture-of-experts AI model for 22 Indian languages. Backed by India AI Mission, it advances sovereign AI with transparency, data provenance and real-world applications in healthcare and governance.
PARAM 2 Launch: BharatGen Powers India’s Sovereign AI Revolution Across 22 Languages
BharatGen unveils PARAM 2, a 17 million-parameter AI model built for 22 Indian languages, advancing India’s linguistic and technological sovereignty. Developed with nine premier academic partners and backed by the India AI Mission, PARAM 2 integrates transparency, provenance, and ethical AI governance. From healthcare to governance, BharatGen’s PARAM initiative marks a defining leap toward India’s trillion-parameter AI future.
Alt: Prof Ganesh Ramakrishnan outlines a bold push to build sovereign AI rooted in Indian languages, data transparency, and domestic capability.
India’s Leap Toward Sovereign AI
In an insightful conversation with G. Ganapathy Subramaniam of The Federal, BharatGen, founder of BharatGen (BharatJain AI Ecosystem Foundation), unveiled the next major step in India’s quest for sovereign artificial intelligence — PARAM 2. This milestone initiative flags India’s entry into large-scale, multilingual AI built entirely around Indian ethics, data, and linguistic roots.
PARAM 2: Designed for India, Built by India
Developed under a collaborative consortium led by IIT Bombay and comprising nine academic institutions, PARAM 2 received over ₹1,235 crore in support from the Department of Science and Technology and the India AI Mission. This cooperative model is distinctly Indian — open, transparent, and driven by shared innovation rather than competition.
Data Provenance and Sovereignty: India’s Ethical Core
BharatGen’s ecosystem positions data provenance — the ability to trace every element of data — as a pillar of ethical AI. By partnering with five certified vendors, BharatGen collects speech data down to the pin code level, ensuring every dataset reflects India’s true linguistic diversity and is fully traceable.
This system aligns with the Prime Minister’s vision of infusing Indian thought and observability into AI development — ensuring that India leads not only in innovation but in responsibility and trust.
Real-World Impact: AI in Hospitals and Governance
Beyond laboratories, BharatGen’s models are already shaping real-world applications. Through collaboration with Amrita Hospitals, the organization introduced Medsum AI, a speech-to-text system that helps doctors generate clinical summaries while maintaining patient consent, privacy, and data ownership.
Meanwhile, Indian state governments are exploring PARAM 2’s language models for public-service AI, linking new policy drafts to historic records, speeding up citizen grievance redressal, and simplifying record access. BharatGen’s approach — co-designing with domain experts — ensures these tools are not just intelligent, but practical and culturally sensitive.
India’s Global Competitiveness in AI
BharatGen and Professor Ganesh Ramakrishnan envision scaling this initiative to 1 trillion-parameter models, positioning India among the world’s most advanced AI nations. As Professor Ganesh noted in the interview, “When India collaborates, one plus one doesn’t make two — it makes eleven.”
Infrastructure and Hardware: The National Backbone
Building AI at this scale requires massive computing muscle — and India is answering that call. BharatGen’s ecosystem collaborates with the National Supercomputing Mission and the National Semiconductor Mission to empower AI workloads.
Strategic partnerships with Larsen & Toubro and Nvidia are bolstering India’s GPU infrastructure, enabling world-class AI training within our borders. BharatGen contributes workload logs and trace data to these missions, ensuring hardware and software innovation stays aligned with India’s sovereign design principles.
Transparency and Ethics: An Indian Blueprint for Trustworthy AI
What distinguishes BharatGen’s PARAM effort is the focus on observability — the technical capability to trace how an AI model learns, decides, and acts. This full-stack accountability safeguards against biases and enables governance in line with India’s ethical and legal values. By embedding transparency, attribution, and clear data lineage, BharatGen is not just building AI models — it’s engineering trust at scale.
The Collaborative Future of Sovereign AI
As the discussion with The Federal revealed, BharatGen’s journey embodies a broader Indian philosophy — technology as a public good. The organization operates as a Section 8 non-profit, ensuring that research and deployment serve citizens first. From healthcare and education to governance and finance, PARAM 2 marks a turning point where India transitions from being an AI consumer to becoming a global producer and innovator.
A New Chapter for India’s AI Renaissance
Through PARAM 2, BharatGen and its partners are not just launching a model — they’re laying the foundation for India’s sovereign AI renaissance built on transparency, collaboration, and purpose. PARAM 2 stands as a testament to what India can achieve when its intelligence — human and artificial — works together for the greater good.
Source: The Federal News
