AI research lab NeoCognition gets $40M in seed funding to develop human-like learning agents

Marketers are actively lobbying AI researchers to develop innovations that will make AI more reliable and effective.

Yu Suthe Ohio State professor who heads the AI ​​lab, said he initially resisted pressure from VCs to sell his work. He finally took the leap last year and started his career at a startup when he saw that improving the model could make agents a reality.

NeoCognition, a startup Su describes as a research lab developing self-learning AI assistants, just came out with $40 million in seed funding. The round was led by Cambium Capital and Walden Catalyst Ventures, with participation from Vista Equity Partners and angels, including Intel CEO Lip-Bu Tan and Databricks co-founder Ion Stoica.

“Today’s agents are very conservative,” Su (pictured right) told TechCrunch. “Every time you ask them to do a job, you’re taking a leap of faith.”

According to Su, the issue lies in instability. Current agents, whether Claude Code, OpenClaw or Perplexity’s computer tools, get the job done right about 50% of the time, he said.

Since agents are still unreliable, they are not ready for reliable, independent operators, Su told TechCrunch. NeoCognition aims to change this by creating an artificial intelligence system that can learn itself to become an expert in any domain, similar to how humans learn.

Su argues that although human intelligence is vast, its true strength is our unique abilities. When we enter a new environment or activity, we become familiar with its rules, relationships, and consequences.

NeoCognition is an architectural aid to demonstrate this process.

“For people, our continuous learning is to create a global model for any job, any location,” Su said. “We believe that the agents to become professionals, they must learn on their own to create a brand for each small country.”

Su sees this opportunity for rapid standardization as the missing link to enabling AI to work reliably on its own.

While it is possible to train assistants for independent tasks, they must be custom-made to be independent. NeoCognition is different because it is a generalist builder that can learn itself and specialize in any field.

NeoCognition wants to sell its systems to businesses, including traditional SaaS companies, who can use them to build a workforce or enhance existing offerings.

Su pointed out that the investment from Vista Equity Partners is very important for this purpose. As one of the largest enterprise computing platforms, Vista is able to give NeoCognition access to more companies looking to improve their business with AI.

NeoCognition currently has about 15 employees, most of whom have PhDs.