Graphics SW/Computer Vision Engineer

In this role, you will work alongside experts from diverse backgrounds to shape the future of on-device AI across mobile platforms (phone, PC, tablet, and automotive), redefining how people interact with the physical and digital world. Your core mission is to research and develop generative and 3D spatial AI that runs efficiently on-device — enabling machines to understand the real world and generate high-quality images, video, and 3D content in real time, under tight compute, memory, and power constraints. What you'll do - Research and develop on-device generative models for image, video, and 3D content creation and editing - Build algorithms for 3D reconstruction, depth estimation, and spatial/visual understanding - Optimize models for edge deployment: quantization, distillation, pruning, and NPU/GPU acceleration - Take ideas from prototype to shipping product, working with hardware, platform, and product teams Qualifications - PhD in Computer Vision, Machine Learning, Computer Graphics, or a related field - Proven ability to turn research concepts into state-of-the-art techniques that hold up against real-world complexity - Strong background in one or more of: diffusion / flow-matching / autoregressive generative models; depth estimation and 3D scene - reconstruction (NeRF, 3D Gaussian Splatting); 3D generative AI; multimodal LLMs and visual understanding; neural networks and transformers - Experience with efficient inference and model compression for resource-constrained hardware is a strong plus - Core programming: primary development in C++ with GLSL/HLSL compute shaders (Vulkan, OpenGL); Python, PyTorch, and CUDA for research and proof-of-concept work - Familiarity with mobile inference stacks (e.g., TensorRT, CoreML, QNN, ExecuTorch, NCNN) is a plus