Physical AI for the Adaptive Factory
From human demonstration to autonomous execution in 60 seconds. The first physics-based imitation engine for industrial arms.
Backed By and Partnering With:
High Mix, Low Volume
90% of manufacturing tasks are too variable for hard-coded automation. Custom engineering costs kill the ROI.
One-Shot Learning
We replaced coding with seeing. Our physics-engine infers assembly logic from a single human demo.
Retrofit, Don’t Replace
Deployed on existing hardware in 60 minutes. Turning legacy fleets into intelligent agents.
Deep Tech Innovation. Industrial Execution.
We aren’t just researchers. We are serial entrepreneurs with 40+ years of manufacturing experience.
Name: Daniele Bernardini
Title: CEO
Bio: Serial entrepreneur and TUM Researcher. Former Software Architect at Siemens & J&J. Leading the convergence of rigid-body physics and deep learning.
Name: Lukas Dirnberger
Title: CTO
Bio: PhD Candidate in ML/Robotics at TUM. Architect of the proprietary Deterministic Imitation Engine and early prototype design.
Name: Francesco Scalise
Title: CRO
Bio: 20-year manufacturing veteran serving clients like Lamborghini & Caterpillar. Expert in industrial sales cycles and go-to-market strategy.
Name: Giorgio Calapà
Title: COO
Bio: MBA with 15 years managing complex projects for Citigroup, Bank of America & Bloomberg. Ensuring operational discipline and financial execution.
We are looking for a Senior Scientist (PhD) to bridge the gap between Perception and Control. First-author publications in ICRA, IROS, CVPR, or equivalent required. View Role & Apply
Contacts
Cognivix Inc.
505 Montgomery St Ste 1100
San Francisco, CA 94111
United States of America
Cognivix s.r.l.
Via Socrate 37
65127 Pescara (PE)
Italy