SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Wayve is building an AI platform for autonomous driving using end-to-end learning from real-world experience. The Simulation team is advancing AV research through generative world models and synthetic data generation.
In this role, you will bridge ML research and engineering to build, scale, and optimize next-generation world model architectures (GAIA and successors) and integrate them into high-throughput generation and training infrastructure. Your work will enable synthetic data to dramatically accelerate autonomy development.
Day-to-day responsibilities include:
- Post-training and ablating GAIA checkpoints for rig transfer and pose transfer, inspecting failures like black margins, odometry bias, flickering, and curvature issues
- Managing generation at scale: launching SDS/Flyte jobs for tens of thousands of segments, debugging GPU capacity, KV cache, DDIM step counts, and data correctness
- Landing synthetic data into training pipelines: binarizing synthetic data into corpus-compatible tables, wiring sampling hooks, training RL models, and measuring on-road impact
- Expanding capabilities to new vehicle platforms and safety-critical scenarios (OEM bring-up, Emergency Lane Keeping, Automatic Emergency Braking)
- Diagnosing and fixing geometry, calibration, and controllability failures that determine whether generated video is training-grade
- Improving throughput and yield through inference optimizations and self-serve workflows
- Collaborating with world-model researchers, infrastructure engineers, and driving-model owners to keep generation, evaluation, and training aligned
You will own a capability end-to-end, not manage a ticket queue. The role sits between research and platform engineering, requiring both technical depth and cross-functional collaboration.
Wayve operates a hybrid model with main hubs in London, Sunnyvale, Yokohama, Herzliya, Vancouver, and Leonberg. The company offers relocation support, visa sponsorship, learning budgets, comprehensive benefits, and meaningful equity.
REQUIREMENTS:
- 4+ years in applied ML or research engineering with a track record of training and shipping neural nets (not just operating data platforms)
- Strong Python and PyTorch (or equivalent); comfort with GPU training, debugging, and reading model code
- Hands-on experience with video, generative, or world models (diffusion, flow-matching, autoregressive video, novel-view synthesis, neural rendering, or similar)
- Working knowledge of cameras and 3D geometry (multi-camera rigs, intrinsics/extrinsics, warps/reprojection) and understanding of how they impact generation and downstream training
- Evidence of taking generated or simulated data into a trained downstream model and measuring impact (mix ratios, ablations, failure analysis)
- Ability to operate generation or training at real scale (multi-GPU jobs, workflow orchestration, large video artifacts) and make that path reliable
- Collaborative, experimental working style with researchers and platform engineers
The company encourages applications from candidates passionate about autonomy even if not meeting every requirement.
