Recently, the artificial intelligence research laboratory NeoCognition, founded by Professor Yu Su from The Ohio State University, officially exited stealth mode and announced a $40 million seed round funding. This round was led by Cambium Capital and Walden Catalyst Ventures, with additional investments from Vista Equity Partners and industry veterans such as Intel CEO Lip-Bu Tan and co-founder of Databricks Ion Stoica.

NeoCognition's core goal is to address the current issue of low reliability in AI agents. Yu Su pointed out that tools such as Claude Code, OpenClaw, or Perplexity currently have success rates of about 50% when performing tasks, and this uncertainty makes it difficult for AI to work independently without human supervision. NeoCognition aims to change the existing "generalist" model and instead develop a general-purpose system with "self-learning" capabilities. This system can build a model of the micro-world like a human expert when entering a specific profession or environment, thus quickly mastering the rules and logic of a particular field.

Differing from traditional AI that requires manual customization for specific vertical fields, NeoCognition builds AI agents capable of self-evolution and expertise in any field. In terms of business strategy, the company plans to primarily target the enterprise market, especially mature SaaS companies, helping them build AI employees or achieve AI modernization of their products. Notably, the involvement of private equity giant Vista Equity Partners is seen as a key strategic move for NeoCognition to gain industry scenarios and connect with enterprise clients.

Currently, the NeoCognition team consists of about 15 people, most of whom hold doctoral degrees. As capital shifts toward startups with fundamental research capabilities, AI agents are moving from simple task execution to reliable productivity tools with deep professional capabilities.