Trajectory raises $45M for AI model training platform
The one-year-old Stanford-founded firm, now valued at $300 million, is opening post-training tools so specialists can improve AI models without research teams.
Trajectory raised $45 million in a Series A round at a $300 million valuation.
Sequoia led the round; Nvidia, Bessemer Venture Partners and Bain Capital Ventures also took part.
The 18-person startup’s AI model training platform helps open-source models improve from real feedback.
Customers already include Harvey, Rogo, Decagon, Mercor and Nvidia; the platform is now publicly available.
Co-founders Ronak Malde, Michael Elabd and Arjun Karanam previously worked at DeepMind and Apple.