NeoCognition raises $40M seed round to build self-learning AI agents

In short: AI startup NeoCognition has raised $40 million to build AI “agents” that can learn new tasks and become more reliable over time.

What happened

NeoCognition, a new startup founded by Ohio State University professor Yu Su, has come out of “stealth,” meaning it was working quietly before making a public announcement.

The company raised $40 million in seed funding, which is early-stage money meant to help a company build its first real product and hire staff. The round was co-led by Cambium Capital and Walden Catalyst Ventures. Other backers include Vista Equity Partners and several individual investors, including Intel CEO Lip-Bu Tan and Databricks co-founder Ion Stoica.

Su told TechCrunch that many current AI agents are not consistent. By “agents,” he means AI tools that can carry out tasks on a computer for you, like a helper that clicks buttons, runs steps, and reports back. Su said today’s agents may only complete tasks as intended about half the time, which makes them hard to trust for important work.

NeoCognition says it is building agents that can keep learning on their own and become specialists, the way a person can learn a new job by understanding its rules and routines (like a new hire who gets better each week).

Why it matters

Companies are interested in AI that can do real work without constant babysitting. If NeoCognition can make these tools more dependable, it could lead to software that handles more day-to-day tasks for businesses, with fewer errors and do-overs.

Source: TechCrunch AI