Former University of Washington AI experts Yejin Choi and Carlos Guestrin have secured $110 million for Noeri, a startup that enables enterprises to fine‑tune models by learning directly from their internal workflows and datasets.
Noeri, the AI startup founded by former University of Washington researchers Yejin Choi and Carlos Guestrin, has closed a $110 million funding round to build technology that lets enterprises fine‑tune large language models using their own data and business processes.
Funding round and investors
The Series B round was led by Andreessen Horowitz with participation from Sequoia Capital, Lightspeed Venture Partners, and several strategic corporate investors. The capital will accelerate Noeri’s product development and expand its go‑to‑market team.
How Noeri’s platform works
Noeri’s core offering lets companies upload proprietary datasets and workflow definitions into a secure environment. The platform then applies a proprietary adaptation layer that aligns a pre‑trained foundation model with the organization’s specific terminology, decision logic, and compliance constraints.
Unlike traditional fine‑tuning, which often requires extensive labeling and model retraining, Noeri’s approach uses reinforcement signals extracted from real‑time process outcomes, enabling rapid iteration without exposing sensitive data to external services.
Target markets and early adopters
The startup is focusing on regulated sectors such as finance, healthcare, and manufacturing, where data privacy and process fidelity are paramount. Early pilots include a major bank that reported a 30% reduction in manual review time after integrating Noeri’s AI into its loan underwriting workflow.
- Financial services – compliance‑aware document analysis
- Healthcare – patient record summarization
- Manufacturing – predictive maintenance recommendations
Challenges and outlook
Noeri must navigate stringent data‑security regulations while delivering performance gains that justify the investment. The founders’ academic pedigree and prior work on large‑scale language models give them credibility, but market adoption will depend on seamless integration with existing enterprise IT stacks.
For readers interested in broader AI developments, the AI category on uxc.news offers daily coverage of emerging technologies and startup funding trends.
Comments
No comments yet.