Mimic Robotics was founded in 2024 as a spin-off from ETH Zurich by Stefan Weirich, Stephan-Daniel Gravert, Elvis Nava and Benedek Forrai, with robotics professor Robert Katzschmann as scientific advisor. The team set out to solve one of automation's hardest problems: the dexterous, fiddly manual tasks that fill factories but have proven impossible to automate with traditional, rigidly programmed robots.
Mimic's approach is rooted in learning from human demonstration. Instead of hand-coding every motion, the company collects demonstrations of people performing tasks directly on the factory floor and trains general-purpose AI models to reproduce that dexterity. The result is a system that can pick up new tasks far faster than conventional automation and handle the variability of real-world parts and materials.
The physical side combines widely available, off-the-shelf robot arms with Mimic's own in-house dexterous robotic hands. By developing the hand hardware itself, the company controls the most critical and difficult component for fine manipulation, while leaning on commodity arms keeps the overall system cost-effective and easier to deploy.
Mimic positions itself as a frontier physical-AI company bringing human-level dexterity to industrial automation, targeting manufacturers struggling with labor shortages for delicate, repetitive manual work. In November 2025 it raised a $16 million seed round led by Elaia and Speedinvest, with Founderful, Sequoia Scout Fund and others participating, bringing total funding above $20 million following an earlier pre-seed. The capital funds expansion of its AI models, dexterous hand development and early industrial deployments as the company works to prove that learned dexterity can finally automate tasks long thought to require human hands.