Research Scientist, Foundational Models

At UVeye, we’re on a mission to redefine vehicle safety and reliability on a global scale. Founded in 2016, we have pioneered the world’s first fully automated suite of vehicle inspection systems, with advanced AI at the core, spanning machine learning, GenAI, and computer vision across the automotive sector.

With close to $400M in funding and strategic partnerships with industry giants such as Amazon, General Motors, Volvo, and CarMax, UVeye stands at the forefront of automotive technological advancement. Our growing global team of over 300 employees is united around a culture of innovation, excellence, and teamwork as we build a safer automotive world.

We are looking for an experienced Research Scientist, Foundational Models to drive UVeye’s long-term foundational model effort. You will design and train a single model that unifies all of UVeye’s inspection segments (cars, trucks, fleets, tires, undercarriage, and future verticals) and serves as the basis for distillation into smaller, cheaper production models.

A day in the life and how you’ll make an impact

• Design and run large-scale pretraining experiments over UVeye’s multi-segment inspection data.
• Develop self-supervised and multi-task learning recipes that transfer cleanly across vehicle segments and use cases.
• Build distillation pipelines that hand off compact, efficient models to production teams.
• Drive fast-adaptation methods so new segments and customers can be onboarded with minimal labels.
• Set and uphold UVeye’s research bar through rigorous evaluation, ablations, and publication-quality work.
  
  

• 4+ years of experience in deep learning research, with depth in foundation models, self-supervised learning, or multi-task learning.
• Track record training large-scale models on real-world data, ideally with multi-modal or vision data.
• Strong Python and PyTorch (or equivalent), and experience with distributed training.
• Experience with model compression, distillation, or efficient inference, an advantage.
• Publications at top AI venues (NeurIPS, ICML, ICLR, CVPR, ICCV, or similar), a strong advantage.
• Comfortable in unexplored territory, independent, and a strong communicator across research and engineering.
• M.Sc./Ph.D. in Computer Science, Electrical Engineering, Data Science, or a related field, a strong advantage.