Trajectory raises $45 million at a $300 million valuation for a platform that lets domain experts post-train AI models

Trajectory, a startup founded about a year ago by three Stanford classmates, has raised a $45 million Series A at a $300 million valuation, TechCrunch reported on November 2. Sequoia led the round, with Nvidia, Bessemer Venture Partners and Bain Capital Ventures taking part. At the same time, the company made its post-training platform publicly available.

What the platform does

Trajectory sells what it describes as a continual-learning system. A business starts from an open-source model, connects the platform to its own applications and gathers feedback on how the model performs in real use. That feedback is then used to retrain the model so it gets better at narrow, specialised work over time.

The pitch is aimed at people who know a field well but do not write training code. Co-founder Ronak Malde told TechCrunch that general foundation models lack the depth of a professional with decades on the job, and that customers should be the ones who define what good output looks like for their organisation. Harvey and Rogo, for example, use the platform to post-train open-source models for legal and financial tasks. Decagon, Mercor and Nvidia are also listed as customers.

Team and funding history

Malde and Michael Elabd previously worked at Google DeepMind, while Arjun Karanam worked on generative foundation models at Apple. The company has 18 employees, according to TechCrunch.

The round was first reported in August. Dealroom, citing The Information, then described a $40 million Series A at a $300 million post-money valuation, led by Sequoia with Nvidia and Bessemer, about two months after a $15 million seed at $115 million. TechCrunch now puts the Series A at $45 million and adds Bain Capital Ventures; it is not clear whether the round grew or the earlier figure was incomplete. Trajectory's own website could not be accessed to confirm the details, and pricing was not disclosed.

A crowded field

Trajectory is entering a busy corner of the market. TechCrunch names Thinking Machines Lab, with its Tinker training service, Applied Compute, valued at $3.25 billion, and Prime Intellect as competitors. All are betting that companies will want to tune open models on their own data and workflows rather than rely only on general-purpose systems from the largest labs. Whether non-engineers can steer that process reliably, without specialist help, remains to be shown at scale.