Enact Intelligence Exits Stealth With $30M Pre-Seed From Mensch and LeCun

Swiss robotics startup Enact Intelligence has raised a $30m pre-seed round — an unusually large sum for the earliest stage of venture funding — with backing from Mistral CEO Arthur Mensch and Meta's former AI chief scientist Yann LeCun.

The company emerged from stealth today. Based in Switzerland, Enact builds the underlying technology for general purpose robots. It's a wager that the next leap in robotics comes not from better cameras, but from better senses of touch.

Enact's approach differs from most robotics companies, which rely on cameras and visual data. Instead, Enact's robots manipulate objects with greater accuracy by sensing force and touch. The target markets are industrial settings and real-world work beyond the factory floor.

Who is putting in the money?

The round was led by institutional investors Redalpine, Notion Capital and Firstminute Capital. The angel list reads like a who's-who of the French and European AI scene: Arthur Mensch, co-founder and CEO of Mistral; Yann LeCun, who served as Meta's chief AI scientist; and Thomas Wolf, co-founder of Hugging Face.

The $30m figure matters because pre-seed rounds typically run far smaller. Enact landed it as investors continue to pour capital into robotics, chasing startups that can make machines genuinely useful in physical environments.

What makes the technology different?

CEO and co-founder Manuel Yves Galliker previously worked at Physical Intelligence, one of the hottest US robotics startups. He frames Enact's edge as a matter of what robots can feel, not just see.

"In particular, our robots sense force and touch through their own actuators, so they feel what's happening in real time, and learn more data-efficiently while simultaneously pushing reliability and performance," Galliker said. "A technology that works across every form factor, built for real-world deployment from day one."

That last point is the pitch in one line: the sensing technology isn't tied to a single robot design. It works across form factors, and it's built to survive outside the lab from the start.

Where does the $30m go?

Enact is deploying the funds on two fronts: expanding the team and accelerating hardware and R&D development. Galliker says the company isn't iterating on one layer of the stack — it's rebuilding all of it.

"We're rethinking the full stack, from model architectures enabling multi-frequency and multi-scale intelligence, to the data pipelines all the way down to the actuators and sensors that feed the model," he said.

That full-stack claim puts Enact in competition with both model-focused robotics labs and hardware specialists. By owning everything from the actuators up to the intelligence layer, the startup is betting that touch-based learning can't be bolted onto systems designed around vision.

The data-efficiency argument may prove decisive. Vision-based robot learning typically demands enormous datasets. If force and touch sensing lets robots learn with less data while improving reliability, Enact's approach could cut both training costs and deployment risk.

With Mensch, LeCun and Wolf on the cap table, Enact also carries credibility with Europe's AI establishment — a signal that the continent's top AI names see physical intelligence as the next frontier, not just language models. The bet now is whether touch-first robotics can move from stealth to the factory floor.