Shodh AI
Category: Foundation Models / LLMs
An AI-for-Science company building India's sovereign foundational models for material discovery, selected under the IndiaAI Mission and collaborating with NVIDIA on Project Skanda. Shodh AI was founded in 2024. The company is led by Dr. Arastu Sharma. Based in New Delhi, India. Team size: 11-50. Total funding raised: $1.1M. Latest round: Grant. Key investors include Government of India (IndiaAI Mission).
- Founded
- 2024
- Headquarters
- New Delhi, India
- Team size
- 11-50
- Total funding
- $1.1M
Value proposition
Compressing material discovery timelines from 10 years to 6 months, and costs from millions to under $1M per new material, using AI foundation models purpose-built for the mesoscale physics of battery cells and energy storage.
Products and solutions
Project Skanda — one of the world's first Mesoscale Foundation Models for Energy Storage, India's first Autonomous Robotic Material Foundry, 7B parameter model for material discovery integrating AI into experimental workflows
Unique value
India's first sovereign scientific brain — a mesoscale foundation model for physical AI that bridges theory and manufacturing, built on NVIDIA Hopper GPUs and PhysicsNeMo, with government backing under the IndiaAI Mission.
Target customer
Battery manufacturers, energy storage companies, material science R&D labs, and enterprises in energy and technology sectors needing accelerated material discovery.
Industries served
Energy Storage, Material Science, Battery Manufacturing, Healthcare (planned), Biotechnology (planned)
Technology advantage
Mesoscale Foundation Model approach (bridging atomic chemistry and manufacturing physics); NVIDIA Hopper GPU + PhysicsNeMo framework integration; India's first Autonomous Robotic Material Foundry for self-driving labs; selected as one of 12 companies under IndiaAI Mission with government compute and funding support; direct recognition from Prime Minister Narendra Modi
How they differentiate
Unlike atomic-scale AI models, Shodh AI targets the mesoscale — the structural architecture inside battery cells that determines life, safety, and charging speed. Combined with autonomous robotic synthesis, it offers an end-to-end AI-to-manufacturing pipeline for material discovery.
Main competitors
Sarvam AI (India, LLMs), Soket AI (India, foundational models), Google DeepMind (global, AI for science), Microsoft Research (global)
Key partnerships
NVIDIA (Hopper GPUs, PhysicsNeMo framework, Project Skanda collaboration), Government of India IndiaAI Mission (selected as one of 12 foundational model companies), Aestr Alpha / Gyanvihar (affiliated university partnership)
Major milestones
Selected as one of 12 foundational model companies under IndiaAI Mission (2025), Launched Project Skanda alpha release at India AI Impact Summit (Feb 2026), Recognized by Prime Minister Narendra Modi for leadership in foundational AI technologies, Announced India's first Autonomous Robotic Material Foundry
Growth metrics
13 employees (as of Jun 30, 2026 per Tracxn); LinkedIn company size 11-50
Market positioning
Positioned as India's sovereign AI-for-science champion, analogous to DeepMind for the UK, with government backing under IndiaAI Mission 2.0 and direct recognition from the Prime Minister. Competing globally in the physical AI / AI-for-science race.
Geographic focus
India (primary); global ambitions
Patents and IP
CEO's prior healthcare patent was acquired by GSK (global pharmaceutical company)
About Dr. Arastu Sharma
PhD in Machine Learning and Applied Physics from University of Cambridge (research in collaboration with Microsoft); experience spans Cambridge, Columbia, and Microsoft Research; serial AI entrepreneur whose healthcare patent was acquired by GSK; also CEO of Aestr Alpha
Latest news about Shodh AI
More Foundation Models / LLMs companies
Official website: https://shodh.ai