Machine Learning Engineer

Wayve Wayve · Robotics · Tokyo, Japan · Product & Delivery

Machine Learning Engineer at Wayve, focused on developing and deploying end-to-end driving models for autonomous vehicles. The role involves leading projects on personalized and collaborative driving, building evaluation pipelines, curating data, and influencing ML system design for production. Requires expertise in deep learning, Python, and production ML systems, with experience in real-time systems or robotics.

What you'd actually do

  1. Develop and improve end-to-end driving models with state-of-the-art performance, robustness, and generalization.
  2. Lead projects on personalized and collaborative driving, including behavior conditioning, comfort tuning, and user alignment.
  3. Build evaluation pipelines and metrics for both closed-loop and open-loop driving performance and product readiness.
  4. Curate and mine real-world and synthetic data to drive scenario diversity, coverage, and feature-specific development.
  5. Influence architecture choices, training methodologies, and deployment pathways for production-scale learning systems.

Skills

Required

  • shipping deep learning systems to production
  • deep learning (esp. sequential models, control, planning, or perception)
  • Python
  • C++
  • CUDA
  • PyTorch
  • software engineering practices
  • real-time systems or robotics
  • simulation- or vehicle-in-the-loop components
  • lead technical initiatives
  • drive alignment
  • mentor engineers

Nice to have

  • autonomous driving
  • imitation learning
  • trajectory prediction
  • personalization
  • human behavior modelling
  • driver intent inference
  • integrating ML systems into production hardware
  • multi-agent simulation

What the JD emphasized

  • shipping deep learning systems to production
  • real-time systems or robotics
  • simulation- or vehicle-in-the-loop components

Other signals

  • develop and improve end-to-end driving models
  • lead projects on personalized and collaborative driving
  • build evaluation pipelines and metrics
  • curate and mine real-world and synthetic data
  • influence architecture choices, training methodologies, and deployment pathways