Senior Research Engineer - Video Foundation Models (Pre - Training)

Accountabilities:

• Design, develop, and scale video foundation models focused on realistic, controllable, and expressive human-centric video generation.


• Build and optimize latent diffusion architectures and conditioning mechanisms for attributes such as pose, emotion, camera control, and script guidance.


• Advance distributed training strategies across multi-GPU and multi-node environments while improving training efficiency and stability.


• Develop and refine evaluation frameworks that combine automated metrics with structured human assessments.


• Optimize inference pipelines to improve latency, scalability, cost efficiency, and output quality for production deployment.


• Conduct rigorous experimentation, ablation studies, and performance analyses to guide model architecture and training decisions.


• Contribute to engineering best practices, including reproducibility, experiment tracking, monitoring, CI/CD, and infrastructure reliability.


• Collaborate closely with cross-functional teams to ensure research outcomes translate into measurable product impact.

Requirements

• Strong experience training deep learning models at scale within production or research environments.


• Advanced proficiency in Python and PyTorch.


• Hands-on experience with diffusion models, particularly within image generation; video diffusion experience is highly desirable.


• Practical expertise in large-scale distributed training using multi-GPU and multi-node systems.


• Solid understanding of distributed learning frameworks such as DDP, FSDP, DeepSpeed, or similar technologies.


• Strong experimental design skills with the ability to interpret complex or noisy results and make data-driven decisions.


• Experience working with modern machine learning infrastructure, including CUDA, AWS, Docker, SLURM, and CI/CD workflows.


• Excellent communication skills with the ability to present technical findings clearly and scientifically.


• Ability to balance research exploration with product-focused execution and delivery.


• Experience with video diffusion models, avatar generation, world models, GANs, VAEs, or production inference optimization is considered a strong advantage.


• Self-driven mindset with the ability to work independently while collaborating effectively within distributed teams.

• Competitive compensation package including salary, bonus opportunities, and stock options.


• Fully remote work flexibility across Europe, with optional hybrid office access in select locations.


• 25 days of annual leave in addition to public holidays.


• Opportunity to work on cutting-edge generative AI technologies with real-world impact.


• High-ownership environment where research directly influences production systems and customer experiences.


• Access to a collaborative and technically exceptional team of AI researchers and engineers.


• Regular team gatherings, planning sessions, and social events.


• Additional location-specific benefits and perks.


• Career growth opportunities within a rapidly expanding AI organization.

This listing is from ats_lever. View original listing ↗