Jonas Andrulis has launched CNTR — a startup designing AI systems that collaborate with humans rather than replacing them.
Jonas Andrulis, the co-founder and former chief executive of Aleph Alpha — once one of Germany’s most closely watched artificial intelligence startups — has revealed details of his next venture, a company called CNTR that is developing AI systems designed to work alongside humans rather than supplant them.
The startup, which launched last month with backing from German consultancy Roland Berger, takes its name from “centaur chess” — a format in which human and computer players compete together as a team. The reference is deliberate. CNTR’s founding premise is that AI systems in industrial environments cannot operate effectively without meaningful human integration, and that the industry’s push toward full automation is producing systems that are both technically fragile and organisationally destructive.
Also Read: Most Boards Were Built for a Pre-AI World. The Bill Is Coming Due.
The company says its technology allows AI agents to ask humans clarifying questions in real time — a mechanism designed to reduce hallucinations and catch the errors that autonomous systems routinely miss when operating without human oversight.
Andrulis was blunt about what motivated the venture. “Most AI systems today are built to replace human labour. Humans are reduced to temporary gap-fillers. That’s a dead end — for the technology itself as well as for the companies putting their most valuable assets at risk: their teams and their culture,” he said. “We’ve founded CNTR so that humans and machines can learn, make decisions, and solve problems together — not competing, but collaborating.”
The startup also announced that Alejandro Molina, an engineer who has worked at Apple, Amazon, and Aleph Alpha, is relocating from the West Coast of the United States to join CNTR as its chief technology officer.
Also Read: Amazon’s Layoffs Aren’t a Warning. They’re a Preview.
Andrulis spent six years building Aleph Alpha into one of the few European companies developing large language models from the ground up. The company subsequently pivoted away from model development toward helping businesses and governments deploy and use AI — a strategic shift that preceded Andrulis’s departure last year.
CNTR represents a different thesis entirely: that the most significant unsolved problem in industrial AI is not capability, but collaboration — and that the companies which figure out how to keep humans genuinely in the loop will outlast those that have tried to remove them from it.
