Continual learning startup Trajectory announced it has raised $45 million in a Series A funding round at a $300 million valuation. The round was led by Sequoia, with participation from Nvidia, Bessemer Venture Partners, and Bain Capital Ventures. Founded by former researchers from Google DeepMind and Apple, the company has grown to 18 employees.
Trajectory’s platform is designed to automate the fine-tuning process for open-source AI models, allowing nontechnical domain experts such as lawyers and doctors to update models based on practical feedback. By connecting directly to business applications, the platform collects performance data and continuously retrains models to specialize in specific tasks without requiring internal research teams or external consultants.
Alongside the funding announcement, Trajectory is making its post-training platform publicly available. Current customers using the system to post-train models include Harvey, Rogo, Decagon, Mercor, and Nvidia. The company emphasizes a human-in-the-loop approach to continual learning to ensure safety and prevent unintended self-improvement outcomes.
Why it matters
Enables nontechnical domain specialists to directly retrain and improve open-source models without hiring dedicated AI engineers.
Reduces reliance on expensive forward-deployed engineers for enterprise fine-tuning and domain adaptation.
Attracts top-tier backing from Sequoia and Nvidia, validating demand for automated post-training tools in specialized markets.
Source: techcrunch.com



