Staff Machine Learning Engineer

NGV Talent works with an early-stage autonomous mobility series A startup focused on developing next-generation vehicle intelligence systems for safe and scalable transportation of goods. We are partnering with our client to support the search for a Staff-level Engineer to join their autonomy team.Key ResponsibilitiesResearch and develop new approaches using deep learning, neural networks, and large-scale foundation models for autonomous driving systems, including perception, prediction, planning, and controlWork across the full machine learning lifecycle, from data analysis and model experimentation to evaluation and performance validationContribute to end-to-end autonomy systems, including mapping, localization, and SLAM-based approachesCollaborate closely with simulation, product, and autonomy engineering teams to integrate ML models into broader system componentsParticipate in cross-functional initiatives spanning perception, planning, and system-level autonomy developmentQualificationsRequired:Advanced degree (Ph.D. or Master’s) in Computer Science, Computer Engineering, Robotics, Mathematics, Physics, or a related fieldStrong foundation in machine learning and/or computer vision, with experience applying modern deep learning methodsHands-on experience with transformer-based architectures and state-of-the-art ML techniquesProficiency with PyTorch, TensorFlow, or similar ML frameworksHands-on experience in 2D/3D object detection, segmentation, and multi-object trackingFamiliarity with modern vision architectures such as DETR, BEVFormer, Vision Transformers (ViT), or similar transformer-based modelsExperience with 3D perception, including BEV-based methods and multi-view geometrySolid background in deep stereo depth estimation or related depth perception techniquesAbility to work independently in a fast-paced environment while collaborating across teamsPreferred:Publications in top-tier ML, robotics, or computer vision conferences (first-author preferred)Experience with generative models, knowledge distillation, or model inference optimization techniques (e.g., TensorRT)