Ola Founder Bhavish Aggarwal To Invest Rs 2,000 Crore In Krutrim, Launches AI Lab

Bhavish Aggarwal, the founder and CEO of Ola, has made headlines with his significant investment in the field of artificial intelligence. He recently announced a commitment of Rs 2,000 crore into his AI startup, Krutrim, with plans to invest an additional Rs 10,000 crore throughout the next year. This step marks a major move towards enhancing the AI landscape in India.

On a social media platform, Aggarwal shared the news that Krutrim has officially launched an AI lab. In his post, he stated, "Announcing the Krutrim AI lab today! While we've been working on AI for a year, today we're releasing our work to the open-source community and also publishing a bunch of technical reports." This initiative aims to bolster AI development tailored specifically for the Indian context.

Focus on Indian AI Development

Krutrim's mission is to improve artificial intelligence applications for Indian languages, address challenges related to data scarcity, and consider cultural nuances that are often overlooked in global AI solutions. Aggarwal emphasized the importance of this localized approach in his announcements.

This investment comes in the wake of the emergence of new players in the AI sector, including the Chinese startup DeepSeek, which introduced its model, DeepSeek-R1. This model is noteworthy as it claims to perform on par with established western AI companies while being considerably more cost-effective.

Partnerships and Future Plans

In a collaborative effort with Nvidia, Aggarwal disclosed plans to implement the GB200 GPU card as part of the Nvidia Blackwell platform. He anticipates that this system will be operational by March and strives to establish it as the largest supercomputer in India by the end of the year. In his post, Aggarwal remarked, "We’re nowhere close to global benchmarks yet but have made good progress in one year. By open-sourcing our models, we hope the entire Indian AI community collaborates to create a world-class Indian AI ecosystem. Learning to walk before we can run is our focus, and we aim to achieve significant milestones within this year."

Additionally, Aggarwal unveiled the forthcoming release of Krutrim 2, an upgraded version of their large language model (LLM). The initial model, Krutrim 1, was launched in January 2024. Other models developed on the foundation of Krutrim 1 include Chitrarth 1, a vision language model for understanding images and documents; Dhwani 1, a speech language model that excels in speech translations; and Vyakhyarth 1, designed for Indic embedding tasks like search and retrieval-augmented generation. They also introduced Krutrim Translate 1 for text-to-text translations.

Global Benchmark Development

The launch was complemented by the presentation of the "BharatBench" global benchmark, aimed at assessing Indic language performance, alongside various technical reports and research papers detailing their findings.

In a notable application of these advancements, Krutrim will provide LLM services for the Kumbh Sah’AI’yak app. This initiative reflects a practical use of the AI developments that aim to serve a broader audience across India.

Investment, AI, Startup