Senior Scientist (Perception & Control)
Pubblicato il 21-08-2026 - Cognivix in Pescara
The Role: Senior Scientist (Perception & Control)
Prima di candidarsi per questo ruolo, la preghiamo di leggere le seguenti informazioni su questa possibilità, riportate di seguito.
You will not be fine-tuning LLMs. You will be building the geometric glue that allows a robot to understand 3D space from a single demonstration. We are looking for aBridge Builder. You need to be fluent in two languages that rarely speak to each other:Deep Learning(Perception) andControl Theory(Action).
What we are looking for:
- - The Bridge:You understandMachine Learning(Transformers, 3D Vision) ANDRobotics(Kinematics, Trajectory Planning, Impedance Control).
- The Pedigree:PhD in Robotics, Computer Vision or Machine Learning.
- The Proof:First-author publications in top-tier venues (ICRA, IROS, CVPR, ICML, T-RO, RA-L or equivalent).
- The Code:You write Python that survives production, not just Jupyter notebooks that run once.
What we are NOT looking for:
- - Pure NLP/LLM specialists.
- Pure Hardware/Mechatronics engineers.
The Offer
- - Role:Senior Scientist / R&D; Lead
- Location:Pescara, Italy (On-site).
Enjoy the Adriatic coast, high quality of life, and lower cost of living compared to Milan or Munich.
- Compensation: €70,000 / Year
This is a top-tier salary for the region, reflecting the seniority of the role.
- Equity:Significant stock option package in an emerging startup.
We are an “A-Team” by design.
You will not be a cog in a giant machine. You will be joining a compact, elite unit of engineers and researchers who are solving the hardest problem in robotics:Generalizability.
- - Our Foundershave 40+ years of industrial execution.
- Our Techwas born at TUM, one of the world’s premier robotics institutes.
- Our Missionis to give eyes and brains to the millions of “blind” industrial robots currently welding and moving metal in the dark.
The Culture: Science meets Scale
We are building a product, not a research paper. However, we believe that the best way to dominate a market is to have the best technology.
- - Publishing:We encourage our team to publish their breakthroughs. We target top-tier venues (ICRA, IROS, RA-L) because peer review keeps us honest and sharp. Your H-Index matters to us.
- Patents:We aggressively protect our core IP. You will be named as an inventor on patents that define the next decade of manufacturing.
- Speed:We don’t spend months on theory. xysqume We code it, we simulate it (Isaac Sim), and then we test it on real hardware immediately.