Noeri raises $110 million as Nvidia, Snowflake and Cisco back cost-efficient AI push
Enterprise demand for cheaper and more adaptable artificial intelligence tools is helping new startups attract major capital as companies reassess how much they spend on AI services. Noeri, founded earlier this year by two Stanford artificial intelligence researchers, is raising $110 million to build software that reviews and improves clients' AI models and coding tools.
Highlights
- Noeri raised $110 million in a funding round led by Menlo Ventures with backing from Nvidia, Snowflake, Cisco, Madrona, Kindred Ventures and NEA.
- Noeri develops an AI tool that analyzes clients' AI models and software, recommending adjustments to improve performance and cost efficiency for enterprises.
- Co-founders Carlos Guestrin and Yejin Choi report their tool enabled older AI models to outperform GPT-5 for scientific research while lowering operational costs.
Funding round supports enterprise AI optimization
As first reported by Bloomberg, Menlo Ventures is leading the $110 million funding round in Noeri, with participation from Madrona Venture Group, Kindred Ventures and NEA. Nvidia Corp., Snowflake Inc. and Cisco Systems Inc. are also investing in the five-month-old startup, while the company's valuation is not disclosed.Noeri is developing what it calls an AI engineer that reviews a client's artificial intelligence models, coding tools and other software, then suggests changes to improve performance and cost efficiency. Those recommendations are explained and passed to a human for approval, with the aim of helping companies tailor AI systems to their operational needs and budgets.
The company is founded by Carlos Guestrin, director of the Stanford AI Lab and a serial entrepreneur, and Yejin Choi, a Stanford professor who received a MacArthur Foundation "genius grant." The founders say the concept grows out of their research work and from efforts to help a microbiology and immunology team at Stanford find lower-cost ways to use OpenAI's models.
Cost pressures shape demand for AI tools
Choi says in an interview that the team built a smaller AI system that learns how to use other AI models, including GPT-4, as part of its research approach. She says the tool improved older models enough that, for their specific scientific research needs, it outperformed GPT-5.That cost-saving pitch may resonate more broadly as businesses increasingly rethink their spending on AI services. Choi previously worked as a distinguished scientist at Nvidia, where she developed a small model designed to use more powerful AI systems to boost performance while reducing costs, a strategy that now sits at the center of Noeri's commercial offering.
As a reminder, we recently covered Dreamteam funding for its AI-native CRM platform challenging Salesforce and HubSpot.