Noeri Aims to Tailor AI to Every Company's Needs

Noeri develops systems capable of progressing from business activity. The startup aims to help companies avoid choosing between the cost of the most effective models and the inferior results of cheaper options.
Choi and Guestrin Bring Their Research Experience
Co-CEOs of Noeri, Yejin Choi and Carlos Guestrin both spent several years at the University of Washington before moving to Stanford. Guestrin contributed to XGBoost, a widely used tool in machine learning, and co-founded GraphLab in Seattle, which became Dato and then Turi. After Turi was acquired by Apple in 2016, he led the group's core teams focused on AI and machine learning.
Awarded the MacArthur Fellowship in 2022, Choi led research on common-sense reasoning at the Allen Institute for AI (Ai2). She left the University of Washington and Ai2 in 2024 for a distinguished scientist position at Nvidia, which she held until founding Noeri in May. She has been teaching at Stanford since early 2025. The company is based in Palo Alto and also has an office in Zurich.
Noeri Proposes to Modify Tools Like Models
Noeri's ambition for continuous improvement is rooted in an experience conducted for a Stanford team in microbiology and immunology, which sought to spend less on using OpenAI's models. According to Choi, the founders designed a small model that learned to leverage older versions of these models. She claims it achieved better results than GPT-5 for the specific research needs of this team, much to the surprise of the founders.
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisissez votre rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Noeri explains that its software examines the company's data and workflows, then suggests adjustments for a specific task. These can pertain to instructions, tools, the choice of model at each step, or even the models themselves. The company states that every proposal is accompanied by an explanation and presented to a person for validation, following the principle of a code review. It announces compatibility with both open-weight models and closed-weight models.
Founders Challenge Dependence on a Single Model
The name Noeri, which means "intelligence, in plural," reflects the founders' refusal to see a single powerful model dominate usage. At the IA40 summit organized by Madrona at the end of September, Guestrin defended the idea that intelligence should not belong to one or two companies that hold the models.
At the launch, Choi and Guestrin likened centralized AI to a single provider that would also advise competitors, could freely revise its terms, and could not be internalized. Regarding ownership, Noeri assures that clients retain their models and processes developed with its software. This guarantee remains subject to the licenses applicable to third-party models.
Menlo Ventures Leads $110 Million Funding Round
Noeri has raised $110 million in a round led by Menlo Ventures, without disclosing its valuation. Madrona, Kindred Ventures, Wing Venture Capital, and NEA, based in Seattle, participated, alongside Cisco Investments, Nvidia, and Snowflake. For Madrona and NEA, this marks a reunion with Guestrin: the funds had led a $6.75 million investment in GraphLab in 2013.
The adaptation of models for businesses also involves other companies. Thinking Machines Lab offers Tinker, an adjustment software that developers can use via API. Applied Compute, for its part, deploys engineers to assist its clients.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.