Decart is seeking a Video Generative AI Research Scientist / Engineer to be a foundational part of our US team, driving the evolution of cutting-edge AI interactive video technologies. This role is pivotal as we scale rapidly, backed by top-tier investors like Sequoia and Benchmark.

What youll do

• Design and train state-of-the-art generative models (diffusion, rectified flow, etc.) for video.
• Develop and optimize our proprietary infrastructure for fast inference and rapid training at scale.
• Collaborate across research and engineering to prototype and ship interactive user experiences, closing the loop from idea to live product.
• Contribute to the overall research agenda and help shape the future of human–pixel interaction.

What Decart is looking for

• 5+ years in Generative AI research positions.
• Knowledge or expertise in CUDA.
• Deep knowledge of modern ML frameworks (e.g., PyTorch) and GPU-accelerated training.
• Proven ability to transition ideas from research to production.
• Experience with multimodal (video / audio) data.
• Proficiency in Python.
• Experience training generative models (diffusion, rectified flow, or similar).
• Strong CS fundamentals; advanced degree (PhD in Computer Science, Electrical Engineering, or related field, or MSc with equivalent industry research experience).
• Experience designing and training state-of-the-art generative models.
• Experience with real-time systems.