Domain-Specific, Low-Compute Indigenous AI Ecosystem
Sovereign, efficient AI built for India's scale and diversity
Contact:
chai@iitmpravartak.org.in
Vision
An India, built, sovereign AI ecosystem enabling every citizen and enterprise to access powerful, affordable, domain, specific intelligence in their own language and context.
Problem
Generic large language models demand prohibitive compute, lack domain depth, and ignore India's linguistic diversity, making AI inaccessible for government, MSMEs, healthcare, and agriculture at scale.
Solution
CHAI develops fine, tuned Small Language Models (SLMs, <14B parameters) optimised for specific sectors, running on low, cost hardware with multilingual support, delivering enterprise, grade AI at a fraction of the cost.
Potential
India's AI market is projected at $17B by 2027. SLM, powered solutions can serve 500M+ users across government, agriculture, healthcare, and education within the next decade, replacing costly generic AI deployments.
Target Beneficiaries
Indian government agencies, PSUs, MSMEs, healthcare providers, agri, tech firms, and educational institutions. Internationally: emerging economies and European organisations pursuing sovereign, sustainable AI ecosystems.
Current Users
AI, assistant deployed and under active use across Indian Government departments, PSUs, and industry partners. Currently self, funded and operational with live user feedback loops.
