
Raviraj Jain
@ravirajjain
Partner @ Lightspeed. Focus on AI, Frontier tech and enterprise software
ID: 20754832
http://lsvp.com/team/raviraj-jain/ 13-02-2009 07:35:51
935 Tweet
2,2K Followers
1,1K Following

A super-humanoid for truck unloading! 📦 What if I told you that there is a robot on the market with two powerful arms and the ability to move? This robot is a recently released Mech from Dexterity Robotics! Unlike humanoid robots, the Mech is purpose-built for industrial

🎤⬇️I could not be more excited to announce Lightspeed’s investment in LightSource LightSource is the only AI-native enterprise procurement platform changing the status quo in strategic and direct sourcing (which is even more timely in a world with a changing tariff


Proud to support Zach Dell Justin Lopas and the Base Power team in their mission to deliver affordable, reliable power nationwide through building America’s next-generation power company 💪🏼 Hit us up if you want to work alongside a world-class team - they are hiring! Sebastian



When we Lightspeed India & Lightspeed backed Sarvam AI in 2023, we were already seeing the AI supply chain fragment across all the layers: chips, LLMs, data & even talent. We were seeing it happen in US and in Europe already. AI is a strategic asset for nations and all large

Congratulations to my friend Liu Jiang for the launch of her fund Sunflower Capital. She’s an exceptional early stage investor and a true supporter of her founders. Looking forward to working on many more startups together Liu!

Welcome to the team Bucky Moore! It’s going to be fun!!! :)


Congratulations to Lightspeed-backed Snorkel AI on their $100M Series D and the launch of two major additions to their platform: the AI Evaluation Suite and the Data Delivery Service, both built to help enterprises confidently move from prototype to production with specialized

What if you could get all the benefits of test-time adaptation of AI models, but 100x faster and essentially for free? Jonas Hübotter of ETH Zurich shows how pre-training 1000 expert models and merging them at runtime achieves exactly that—with nearly zero overhead.

