Trajectory raises $45M to make AI model training accessible to nontechnical experts
Reported by TechCrunch, ecosistemastartup and techbuzz3 sources ↓
The one-year-old startup raised $45 million in a Series A round led by Sequoia, valuing it at $300 million.
Key points
- Nvidia, Bessemer Venture Partners and Bain Capital Ventures also joined the round
- Trajectory says its platform is now publicly available
- Harvey and Rogo use the platform to post-train open-source models
- The 18-person startup also counts Decagon, Mercor and Nvidia among its customers
Sources
Details
Trajectory is developing a platform that gathers feedback on how AI models perform in specialized tasks and uses it to retrain them. TechCrunch reports that businesses can connect an open-source model to their applications, allowing specialists to help improve it without relying on an internal research team or outside consultants.
The company says users decide what quality means for their organization and use that feedback to improve models. Examples at Harvey and Rogo include learning from contract reviews and company valuations. Trajectory identifies Thinking Machines Lab, Applied Compute and Prime Intellect as competitors in AI post-training or related tools.
Written from coverage as of
Images
- TechCrunch
- ecosistemastartup
- techbuzz