Key Takeaways
- • Noeri, a Palo Alto-based AI lab co-founded by Stanford computer science professors Yejin Choi and Carlos Guestrin, announced $110 million in funding led by Menlo Ventures, with participation from Madrona Venture Group, Kindred Ventures, Wing Venture Capital and NEA.
- • Strategic investors Cisco Investments, NVIDIA and Snowflake are also participating in the round, according to the company's Business Wire announcement.
- • Noeri is building self-improving AI systems that learn from an enterprise's own data and workflows, with every proposed change sent to a human for approval, and customers retain ownership of the data, workflows and models they develop.
- • The company is positioning its approach as a way to move past the trade-off between paying for the most capable general-purpose model and accepting weaker results from a cheaper one for company-specific tasks.
- • Noeri is also pursuing AI-driven scientific discovery, with its systems forming hypotheses and running experiments on problems such as molecules, proteins and complex technical systems.
Summary
The company's core thesis is that today's AI models are static after training, whereas Noeri aims to build systems that self-organize and improve through on-the-job interactions. According to the company, large enterprises are spending heavily on AI but often struggle to convert that spending into results, because general-purpose models are optimized for breadth rather than company-specific tasks. Noeri's system studies an enterprise's data and workflows and then refines the surrounding components of a task, including the harness, skills, instructions, tools, and the choice of model for each step, and in some cases the underlying model parameters. Each change is explained and submitted for human approval, similar to a code review. The system works with both closed and open-weight models and adapts as new models become available.
Noeri also extends this approach to scientific discovery, where its AI is designed to form hypotheses, run experiments and learn from results, acting as a co-discoverer alongside scientists and engineers. The company argues that frontier-model economics become prohibitive in open-ended discovery work because search consumes far more compute, and that smaller specialized models make sustained discovery affordable. The company states that its technology draws on the founders' prior research, including Choi and Guestrin's joint work on test-time training, Guestrin's self-distillation work (SDPO), and Choi's research on agentic orchestration and pluralistic alignment.
Venky Ganesan, Partner at Menlo Ventures, said the firm moved quickly to lead the round because the team combines world-class research with deployment inside companies. Choi is a MacArthur Fellow named to the TIME100 Most Influential People in AI in 2023 and 2025, and previously was a professor at the University of Washington and a senior director at the Allen Institute for AI. Guestrin co-founded Turi, acquired by Apple in 2016, and subsequently built and led Apple's central machine learning and AI organization; he also co-created the open-source XGBoost project. Noeri says it is working with strategic partners across industries, including some of its investors.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.