Krutrim | UPSC

Krutrim mind map
Recent News
India's first AI unicorn
Valued at $1 billion
Raised $50 million funding
Founded by Bhavish Aggarwal
Plans for AI chips manufacturing
When
Founded last year
Beta version release next month
Full public access in January 2024
Developer APIs in February 2024
Why
To empower India's AI revolution
Addresses lack of computing penetration in India
Aims for AI-first economy
What
Large Language Model (LLM)
Trained on local Indian languages
Understands 22 languages
Generates text in 10 languages
Over 2 trillion tokens
Outperforms GPT-4 in Indic languages
Krutrim Pro
Multi-modal
Advanced problem-solving
Set to launch next quarter
Where
Based in India and the Bay Area, US
Who
Bhavish Aggarwal
Co-founder and CEO of OLA
Visionary behind Krutrim
Matrix Partners India
Led the funding round
How
Building in-house AI chips
Utilizing Indian languages and data
Collaborative efforts with OLA companies
Significance
First Indian AI model to reach unicorn status
Inclusivity in technology
Potential for personalized learning and customer service
Claims to surpass OpenAI's GPT-4 in Indian languages
Challenges
High cost of LLM training
Potential biases and ethical concerns
Competing with established AI models
Way Forward
Addressing AI development challenges
Expanding to multimodal capabilities
Leveraging Indian academic and startup communities
Infrastructure and silicon development for AI

Krutrim, an AI startup founded by Bhavish Aggarwal, has emerged as India’s first AI unicorn, achieving a valuation of $1 billion with a $50 million funding round. It represents a significant step for India’s participation in the global AI landscape, aiming to revolutionize the AI market by focusing on local Indian languages and data. Krutrim’s approach includes developing a large language model that understands multiple Indian languages, plans for an AI chip manufacturing capability, and creating a multimodal advanced version named Krutrim Pro. The venture signifies India’s ambition to be an AI-first economy, addressing the existing gap in computing penetration and aiming to make technology inclusive and accessible across diverse Indian languages.