The fresh capital sets the one-year-old startup's valuation at $300 million as it seeks to automate the complex post-training process for open-source AI models. By keeping human domain specialists in the feedback loop, the company aims to offer a safer alternative to autonomous recursive self-improvement while cutting out costly research consultants. Trajectory's platform works by linking open-source models to live applications, gathering feedback from domain experts like doctors and lawyers to drive continual retraining. This newly public product is already deployed by enterprise clients including Nvidia, Harvey, and Decagon.
Key points
Nontechnical professionals can directly train and refine specialized AI models using perf…: Nontechnical professionals can directly train and refine specialized AI models using performance feedback collected from real-world applications.
Maintaining human subject-matter experts in the feedback loop provides a safer continual…: Maintaining human subject-matter experts in the feedback loop provides a safer continual learning mechanism than fully autonomous self-improvement.
The startup faces active market competition from rival AI firms including Thinking Machin…: The startup faces active market competition from rival AI firms including Thinking Machines Lab, Prime Intellect, and Applied Compute.
Sources checked by our analysis pipeline: 1
Whether developed by OpenAI, Anthropic, or built on open-source alternatives, current AI models share a critical limitation: they don’t learn from their mistakes without further training.
Businesses adapting these models for specialized fields, such as law and finance, typically get around this by fine-tuning general models. However, that often requires hiring internal research teams or relying on expensive consultants, commonly known as forward-deployed engineers (FDEs).
Trajectory, a startup founded about a year ago by three Stanford classmates with experience at Google DeepMind and Apple, wants to automate that process, allowing AI models to learn from feedback and continuously improve at specialized tasks.
The 18-person company just announced that it raised $45 million in a Series A round at a $300 million valuation, led by Sequoia, with participation from Nvidia, Bessemer Venture Partners, and Bain Capital Ventures.
Trajectory co-founder Ronak Malde compared foundational models to a very smart PhD student who’s always starting a new job. “But what the economy values is someone who worked on the job for 30 years,” he told TechCrunch.
Silicon Valley leader Garry Tan has embedded automated systems across his executive duties and daily routines to streamline operations and coding. By feeding decades of private records and organizational history into these workflows, he is testing the limits of productivity augmentation. However, he warns that steep infrastructure expenses currently restrict such powerful, self-directed systems to well-funded elites.
The artificial intelligence company has recruited experienced defense leadership to strengthen its ties with government and defense sectors. Bringing former high-ranking military officials into leadership positions highlights the increasing strategic intersection between advanced technology development and federal security operations. This transition underscores how industry leaders are positioning themselves to address complex national security and regulatory demands.
Venture capital firm Sequoia has backed an artificial intelligence startup focused on expanding accessibility beyond specialized developers. The fresh capital will support software that simplifies the post-training process of machine learning models for users without coding expertise. This substantial early-stage valuation highlights strong investor appetite for commercial tools that streamline enterprise AI adoption.