Staff Tech Lead, Perception
Core
Train multi-billion-parameter spatio-temporal foundation models on massive datasets and deliver pre-trained models for real-time production systems.
Role type
Staff Tech Lead, Perception
Builds
Pre-trained foundation models for real-time systems
Domain
Computer Vision / Deep Learning
Deliverable
production ML models
Required skills
multi-billion-parameter model training, self-supervised learning, image/video encoding, 3D point cloud processing, spatio-temporal reconstruction, vision-language contrastive learning, multi-teacher distillation, dataset curation, Python, Pytorch, JAX
Preferred skills
C++, real-time system design, publications at top-tier conferences (CVPR, ICCV, ECCV, ICLR, ICML, ICRA, RSS, NeurIPS, AAAI, IJCV, PAMI)
Responsibilities
Develop self-supervised learning methods for image/video encoders, 3D point cloud encoders, spatio-temporal reconstruction, and future prediction; Develop multi-teacher distillation recipes; Evaluate latent representation quality and collaborate with post-training teams; Contribute to technical roadmaps
Seniority
Staff, hands-on IC with strategic contribution
