Noeri is building software that studies a company's data, workflows and business processes, then proposes changes to specific tasks to make them more efficient. Suggestions can involve different instructions, different tools, or a different model for a workflow step. Each one comes with an explanation and goes to a human reviewer before it is accepted, much as AI-generated code changes are reviewed.
The system works with both open-weight and proprietary models, and customers keep full ownership of their workflows and the AI systems built for them. Choi told Bloomberg the company targets organizations that must either pay fees they cannot afford for the most capable models or accept results from a weaker one. The idea started with work to help a Stanford microbiology and immunology team use OpenAI models more cheaply.
Menlo Ventures led the round, with participation from Madrona Ventures, Kindred Ventures, Wing Venture Capital, NEA, Snowflake and Cisco Ventures. The startup says it is already working with partners across several industries, including some of its investors.
Why it matters if you build agents
Teams building agents on proprietary models can now evaluate a system that moves workflow steps between models, but every change it proposes still needs human sign-off. Developers comparing vendors should weigh the stated focus on model choice and workflow ownership. The announcement does not cover pricing or general availability, so check access before planning around it.
Get stories like this every weekday
One short email with the day's agent tooling news.