Research & Insights

AI Notes

Research-driven notes on AI strategy, policy, and the technology landscape — written to encourage informed discussion across India’s AI ecosystem.

India AI JourneyIssue #03

Sarvam Epoch 2026: Models, Infrastructure and Applications

On 30 July 2026, Sarvam AI announced across every layer of the AI stack on a single day — models, infrastructure, harnesses and applications. This note documents all of it: the upgraded 105B and its cost router, Saaras V4 and the world's first real multi-speaker Indic ASR, the India-hosted token factory, the six-product Indus platform, and a trillion-parameter frontier commitment defended with arithmetic rather than adjectives. Every quantitative claim is collected in one scorecard, clearly marked as Sarvam's own and not independently verified.

Bhaveshkumar Choithram Dharmani1 August 202634 min read
India AISarvam AISarvam EpochFrontier ModelsAI InfrastructureAgentic AIIndic LanguagesToken SovereigntyIndia AI Journey
Read the note
India AI JourneyIssue #02

India's AI Mission: The Seven Pillars and What Each One Actually Funds

India's AI Mission is usually summarised as one number: ₹10,372 crore. That hides what actually matters to a builder — that its compute portal, its AIKosh dataset platform, and its skilling programmes are live and free to register for today. This note goes pillar by pillar, names all twelve funded model teams (not just Sarvam), and gives the direct links and steps to access each facility yourself.

Bhaveshkumar Choithram Dharmani10 July 202634 min read
India AIIndiaAI MissionAI PolicyAIKoshIndiaAI ComputeSarvam AIBharatGenIndia AI Journey
Read the note
India AI JourneyIssue #01

Where Does India Stand in the Global AI Race? An 8-Layer Framework for Understanding India's AI Strategy

Understanding India's AI position requires looking beyond individual breakthroughs. This note examines India across eight layers — from energy and silicon at the base to governance at the top — and finds genuine, specific strengths alongside real but addressable gaps.

Bhaveshkumar Choithram Dharmani1 June 202635 min read
India AIAI StrategyAI PolicyIndiaAI MissionAI StackSovereign AI
Read the note