Noeri Launches With $110M to Build Self-Improving AI for Companies

Noeri Launches With $110M to Build Self-Improving AI for Companies

Noeri, a new AI lab started by Stanford professors Yejin Choi and Carlos Guestrin, launched on Monday with $110 million in funding to build self-improving AI that companies can own. Menlo Ventures led the round, and Nvidia, Cisco Investments and Snowflake joined as strategic investors. The pitch is simple: instead of one giant model that stops learning after training, Noeri wants many smaller AI systems that keep getting better on the job.

Inside Noeri’s $110 Million Funding Round

The round was led by Menlo Ventures. Madrona Venture Group, Kindred Ventures, Wing Venture Capital and NEA also took part. Three big tech names came in as strategic backers, which hints at where Noeri may find early partners.

DetailWhat We Know
Amount raised$110 million
Lead investorMenlo Ventures
Other investorsMadrona, Kindred Ventures, Wing Venture Capital, NEA
Strategic investorsCisco Investments, Nvidia, Snowflake
FoundersYejin Choi and Carlos Guestrin, co-CEOs
LocationsPalo Alto, with an office in Zurich
Founded2026

The company did not share its valuation. It says it is hiring researchers and engineers in both Palo Alto and Zurich.

What Self-Improving AI Means at Noeri

Most AI models today are frozen once training ends. Whatever a model knew on its last training day is all it knows while it works. Noeri says its systems are different because they learn from the work they do and grow more capable over time.

The company’s AI studies a business’s own data and workflows. Then it tries to improve everything around a task, not just the model. According to Noeri, that can include:

  • The harness, instructions and skills an AI system uses
  • Which tools it calls and which model handles each step
  • When it pays off, the models themselves and their parameters
  • Support for both closed models and open-weight models

Every change is explained and sent to a person for approval, a bit like a code review. Noeri says customers keep ownership of their data, workflows and the models they build, subject to third-party model licenses.

The Cost Problem Noeri Wants to Fix

Noeri argues that companies face a frustrating choice today. They either pay more for the most capable model or accept weaker results from a cheaper one. The company’s bet is that a specialized system tuned for one job can beat that trade-off.

Guestrin put it plainly in the announcement. “We think specialization can move the whole curve,” he said.

The Founders Behind Noeri’s Self-Improving AI

Both founders are well known in AI research. Choi is a MacArthur Fellow who was named to the TIME100 list of influential people in AI in 2023 and 2025. She previously taught at the University of Washington and was a senior director at the Allen Institute for AI.

Guestrin co-founded Turi, the machine learning platform Apple bought in 2016, and later led Apple’s central machine learning and AI group. He also co-created XGBoost, one of the most widely used machine learning libraries. The two worked together at the University of Washington before both moved to Stanford.

The Research Behind the Startup

Noeri says its technology comes straight from the founders’ academic work. That includes test-time training, where a model keeps learning after deployment, and a self-distillation method that lets a model learn from its own mistakes. Choi’s research on agentic orchestration showed that a small model directing a team of other models and tools can beat frontier models at a fraction of the cost.

The name Noeri means “intelligence, in the plural.” It reflects the founders’ view that the future should hold many AI systems of different sizes, not one dominant model.

Science Is Part of the Plan Too

Noeri is not only aiming at business workflows. The company says the same learning machinery can help scientists form hypotheses, run experiments and learn from the results. It mentions work on new molecules, proteins and complex technical systems.

Noeri also points out that open-ended discovery burns through far more compute than answering one question. Smaller, specialized models could make that kind of long search cheaper. That fits a wider shift we have seen in AI research tools toward focused, task-specific systems.

What This Means for You

If you run AI projects at work, Noeri’s launch is another sign that the “one big model for everything” era may be fading. Here is what to keep in mind:

  • Specialized AI systems tuned on your own data may soon compete with top general models on cost and quality.
  • Ownership matters. Ask any vendor who owns the workflows and fine-tuned models you build with them.
  • Human approval for AI changes is becoming a selling point, not a burden.
  • Smaller models directing tools and agents are a trend worth watching as you plan budgets.

Noeri is still early and has not shared customer results, so it is too soon to judge the product. For now, it is worth tracking alongside other business AI tools and the growing wave of enterprise AI agents. You can follow more funding rounds like this on our AI news page.

Sources

Source: Noeri’s official funding announcement (Business Wire release, via Yahoo Finance).

Conclusion

Noeri enters the crowded AI lab race with $110 million, two respected founders and a clear idea: AI that keeps learning after it ships and belongs to the companies that use it. Backing from Nvidia, Cisco and Snowflake gives it strong partners from day one. The big question is whether self-improving systems can deliver real results inside businesses, not just in research papers. If they can, cheaper specialized AI could become a serious alternative to paying top dollar for frontier models.

Newsletter

New AI tools in your inbox

The best new tools and the AI news that matters, in one short email. Unsubscribe any time.

Free · No spam · Unsubscribe any time

✓ Looks like you’re subscribed

Loading Google’s sign-up form…

Looks like you’re already subscribed

Thanks for reading! New AI tools and the AI news that matters will land in your inbox.

Not subscribed yet? Refresh the page to load the sign-up form.

The sign-up form couldn’t load. If you use an ad or content blocker, allow news.google.com and reload the page.

Free instant check before you pay. Takes a few seconds.

Scroll to Top