Trajectory raises $45M to make AI model training accessible to nontechnical experts
Trajectory raised $45 million in a Series A round to automate AI post-training, letting nontechnical experts like lawyers update models.Summary
Trajectory, a startup founded by three Stanford classmates with experience at Google DeepMind and Apple, announced a $45 million Series A round at a $300 million valuation, led by Sequoia with participation from Nvidia, Bessemer Venture Partners, and Bain Capital Ventures. The company addresses a key limitation of current AI models—that they don't learn from mistakes without further training—by automating post-training through continual learning. Businesses connect the platform to their applications so it can collect performance feedback and retrain open source models to improve at specialized tasks. Customers including Harvey, Rogo, Decagon, Mercor, and Nvidia use the platform to help models perform like experienced lawyers or financial analysts, and the platform is now publicly available. Co-founders Ronak Malde, Michael Elabd, and Arjun Karanam argue that keeping humans involved makes their approach safer than autonomous self-improvement, and Malde cites the team's pedigree as its biggest advantage over competitors such as Thinking Machines Lab, Applied Compute, and Prime Intellect.
(Source：TechCrunch)