River AI, founded by xAI co-founder Igor Babuschkin, has raised $1.1 billion in funding led by General Catalyst and AMP PBC. The round includes strategic investments from NVIDIA and AMD Ventures, with additional backing from Y Combinator and Temasek. The Palo Alto-based company offers an API that enables enterprises to complete complex reinforcement learning training runs in 15 to 20 minutes without requiring an infrastructure team, at two to four times the cost savings versus closed-source alternatives. River AI's platform provides state-of-the-art LoRA fine-tuning and reinforcement learning for frontier open weight models. The company aims to make AI ownership accessible to developers and organisations rather than relying solely on general-purpose models. Babuschkin previously worked at Google DeepMind and OpenAI before co-founding xAI.

Y Combinator president Garry Tan said AI-focused startups are achieving unprecedented growth, with some reaching profitability within eight months. Speaking at the incubator's Startup School 2026 event, Tan explained that entrepreneurs can now leverage "personal AGI" — AI agents that work across ventures, run independently, and compound knowledge over time. During recent Y Combinator cycles, most companies used AI agent models. Emergent, from the summer 2024 batch, reached nine-figure revenue in eight months with just 15 people at $15 million annualised revenue. Retell, launched winter 2024, hit $60 million annualised with 40 people. "Before you ever incorporate anything, you can already be running an organisation of one plus your agents," Tan said. He noted such rapid revenue per person was previously impossible in any industry.

Discover how Singaporean poly grads Jamie Yau and Kazel Koh's start-up Collar AI achieved a rare acceptance into the prestigious Y Combinator accelerator. Read more at straitstimes.com. Read more at straitstimes.com.