# Trajectory Lands $45M From Sequoia to Let Lawyers and Doctors Retrain AI Models Without Engineers
> Trajectory raises $45M Series A at a $300M valuation led by Sequoia to automate AI post-training for nontechnical domain experts like lawyers and doctors.
- **Author:** Swati k jagtap
- **Published:** 2026-11-03
- **Updated:** 2026-11-03
- **Category:** Startups
- **Editorial type:** news
- **Topics:** series-a, sequoia-capital, nvidia, bessemer-venture-partners, ai-startups
- **Canonical URL:** https://www.todaysstartupnews.com/startups/trajectory-raises-45m-series-a-sequoia-ai-model-post-training-platform
## Key takeaways
- Trajectory has raised $45 million in a Series A round at a $300 million valuation, led by Sequoia with Nvidia, Bessemer Venture Partners and Bain Capital Ventures participating
- The platform lets businesses start from an open-source model, connect it to their applications, collect feedback on real use and retrain the model automatically
- Harvey and Rogo already use it, and other customers include Decagon, Mercor and Nvidia
- The 18-person company, founded by three Stanford classmates with backgrounds at Google DeepMind and Apple, is opening its post-training platform to the public
Trajectory, a startup that automates the post-training of AI models, has raised $45 million in a Series A round at a $300 million valuation. Sequoia led the round, with Nvidia, Bessemer Venture Partners and Bain Capital Ventures also participating. The company is using the announcement to open its post-training platform to the public.

Trajectory was founded by three Stanford classmates, Ronak Malde, Michael Elabd and Arjun Karanam. Malde is the chief executive. He was among the researchers Google DeepMind brought in when it hired part of the Windsurf team, and Elabd also worked at DeepMind. Karanam previously worked on generative foundation models at Apple. The company has 18 employees and has also hired researchers and product designers from OpenAI, Stripe and Netflix.

The round follows a $15 million seed round announced in May 2026, when Trajectory introduced itself as a research lab and product company focused on continual learning.

The problem Trajectory targets is a limitation shared by current AI models, whether they come from OpenAI, Anthropic or open-source projects. Models do not learn from their mistakes unless they go through further training. Businesses that adapt models for specialised fields such as law and finance usually fine-tune general models, which tends to require internal research teams or expensive outside consultants known as forward-deployed engineers.

Trajectory aims to automate much of that workflow. A business starts with an open-source model and connects the Trajectory platform to its own applications. The platform collects feedback on how the model performs in real use and then retrains the model with that information, so it improves at specialised tasks over time.

Malde compared foundation models to a very capable PhD student who is always starting a new job, while the economy rewards someone with decades of experience in the role. According to him, the platform lets lawyers, doctors and other specialists improve models almost automatically, which could make specialised AI deployments faster and cheaper. Users decide what good output looks like for their organisation, and the platform converts that judgement into model improvements.

Harvey and Rogo already use the platform to post-train open-source models so that they behave more like experienced lawyers or financial analysts. The models learn from previous tasks, such as reviewing a particular type of legal contract or valuing a company with a specific methodology. Other customers include Decagon, Mercor and Nvidia.

Trajectory describes its approach as continual learning, which shares some features with recursive self-improvement, the idea that AI systems can improve their own capabilities. Elabd said that keeping humans involved in the training process makes the approach safer, since fully autonomous self-improvement could produce unintended or even dangerous outcomes.

Interface design is also a focus. Karanam said the company built an interface intended to make post-training approachable for customers who do not have a technical background.

Trajectory is not alone in trying to automate post-training. The company lists Thinking Machines Lab, founded by former OpenAI chief technology officer Mira Murati and the maker of the fine-tuning platform Tinker, as a direct competitor. Other names in the field include Applied Compute, which was valued at $3.25 billion last month, and Prime Intellect, which helps companies build and train their own AI agents. Malde named the team as Trajectory's biggest advantage.

With the platform now publicly available, Trajectory is moving from working with a limited set of design partners to a self-serve product for organisations that want to customise models using their own production data and expert feedback.
## Sources
- [trajectory.ai](https://www.trajectory.ai)