Investors are actively pursuing AI researchers to develop startups focused on improving AI’s reliability and efficiency.
Yu Su, an Ohio State professor leading an AI agent lab, initially hesitated to commercialize his work despite pressure from venture capitalists. However, he launched a startup last year after recognizing how advancements in foundational models could enable truly personalized agents.
NeoCognition, a startup described by Su as a research lab creating self-learning AI agents, has recently emerged from stealth with $40 million in seed funding. The funding round was co-led by Cambium Capital and Walden Catalyst Ventures, with participation from Vista Equity Partners and notable angels like Intel CEO Lip-Bu Tan and Databricks co-founder Ion Stoica.
“Today’s agents are generalists,” Su told TechCrunch, highlighting a lack of consistency with current agents, which accurately complete tasks only about 50% of the time.
Due to this unreliability, agents are not yet trusted as independent workers. NeoCognition aims to resolve this by creating an agent system capable of self-learning to become an expert in any field, akin to human learning.
Su argues that while human intelligence is wide-ranging, its true strength lies in specialization. Humans can quickly adapt to new environments or professions, mastering their specific rules and dynamics.
NeoCognition plans to apply this concept to their agents.
“Our ongoing learning process is building a world model for any profession, any environment,” Su said. “For agents to become experts, they need to learn autonomously to create a model of any particular micro world.”
Su believes this rapid specialization is crucial for AI to function independently and reliably.
While agents can be tailored for autonomous tasks, they often require custom engineering for specific sectors. NeoCognition differs by developing generalist agents capable of self-learning and specialization across various domains.
NeoCognition plans to offer its agent systems mainly to enterprises, including established SaaS companies, to build agent-workers or enhance existing products.
Su emphasized that Vista Equity Partners’ investment is particularly beneficial. As a leading private equity firm in the software industry, Vista can connect NeoCognition with numerous companies eager to integrate AI into their products.
NeoCognition currently employs about 15 staff members, most of whom possess PhDs.