Senior Machine Learning Engineer, ADAS

Overview

In this role you will advance ADAS perception capabilities by developing and refining CV/3D models, working across data, training, evaluation, and iteration. You will contribute to end-to-end perception that can drive in-vehicle decisions, balancing online (on-car) and offline (large-scale) workflows. You’ll collaborate with cross-functional teams to target performance gaps, build scalable data pipelines, and accelerate model development with auto-labeling and pseudo-labeling. This is a hands-on, impact-driven position in a fast-moving autonomous driving company.

Pay / Benefits

• market-based salaries
• equity
• relocation support and visa sponsorship
• hybrid working
• learning and development budget
• health insurance and dental care

Responsibilities

• Train, debug, and improve CV and 3D perception models based on evaluation signals
• Work across the full ML lifecycle: data, training, evaluation, and iteration
• Collaborate with team to prioritize work based on system underperformance
• Build scalable data pipelines including auto-labelling and pseudo-labelling
• Develop core perception capabilities (detection, classification, instance segmentation) across lanes, objects, signs, and lights
• Pursue end-to-end approaches enabling models to act directly in the world without hand-coded intermediates
• Implement online perception that runs fast in-car and offline perception using additional inputs such as LiDAR, maps, and extra compute
• Create the data flywheel by using offline models and 3D tracking/reconstruction to auto-generate labels for faster iteration

Key requirements

• Built and shipped CV-focused deep learning systems
• Experience with 3D perception concepts or pipelines (LiDAR, multi-view geometry, tracking, 3D reconstruction)
• Comfort owning work end-to-end including evaluation and large-scale dataset generation
• Pragmatic problem-solving under real product constraints
• Automotive domain experience is a bonus but not required

• pragmatic problem-solving
• ownership
• collaboration across teams
• CV-focused deep learning systems
• 3D perception concepts (LiDAR, tracking, 3D reconstruction)
• data pipelines and scalable dataset generation

…