Trajectory raises $45M for AI model self-improvement platform — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Trajectory raised $45 million in a Series A round at a $300 million valuation, led by Sequoia with participation from Nvidia, Bessemer Venture Partners, and Bain Capital Ventures
- Trajectory launched a publicly available post-training platform that uses continual learning to help AI models improve over time based on user feedback in specialized domains like law and finance
- Trajectory’s co-founders—Ronak Malde, Michael Elabd, and Arjun Karanam—bring experience from Google DeepMind, Apple, and acqui-hired research teams, and emphasize human-in-the-loop training to avoid risks of autonomous AI self-improvement
- Companies including Harvey, Rogo, Decagon, Mercor, and Nvidia are already using Trajectory’s platform to post-train open-source models for tasks like legal contract review and financial valuation
- Arjun Karanam highlighted that Trajectory’s interface design makes AI post-training accessible to nontechnical experts, a key differentiator in democratizing specialized model improvement
Why it matters: Specialists in high-stakes fields can now refine AI models without relying on costly data science teams, reducing deployment barriers by directly embedding domain expertise into models. At $300M valuation with Nvidia and Sequoia backing, this shift could accelerate adoption of adaptive AI in industries where accuracy evolves with use.
Ask SkimNews
