Applied Scientist/Machine Learning Engineer, Gaia

Start free →

Or apply yourself on the official page →

About the role

Before the detail, here's the challenge you'd help us solve. We build the embodied intelligence that moves real vehicles safely, and the ecosystem a billion machines will run on in the future. Very few people in AI can say this. Every role here, whatever the team, plugs into that. Here’s what this particular role covers. 🛠️ About our Simulation Teams Generative Simulation is the team advancing our end-to-end autonomous driving research, incubating and investing in new ideas that can become game-changing technological advances for Wayve. As a Machine Learning Engineer in the team, you will play a key role in developing next-generation world models and planners, such as GAIA, that can simulate complex, diverse, and temporally consistent driving environments to power faster training, broader testing, and scalable deployment, even in areas and scenarios we have never driven in before. As we push toward the next generation of GAIA, efficiency and interactivity are a major focus: models must run thousands of roll-outs per second, support closed-loop agent interaction, and fit within practical compute budgets. You will work at the intersection of machine learning research, multi-modal modelling, and real-world deployment, tackling questions such as how we can deploy autonomous vehicles in a new geography without collecting any real-world data, and whether synthetically generated environments can fully replace physical testing and data collection.   🧠 Your day-to-day • Setting technical direction: Set technical direction and designing key components of the system. • Deep work / coding: You spend a focused block implementing and debugging new modeling capabilities. • Cross-team collaboration, brainstorming and pairing: you brainstorm ideas with your team mates or pair to solve a difficult problem • Mentorship / async review: You review a teammate’s pull request or experimental design and outcomes and provide feedback • Sync-ups: Share findings with the rest of your team leading to a discussion on designing the next experiments. You align on shared priorities for the week and blockers. • Paper reading/ reading groups: Read one of the latest papers and summarise them for yourself or the team to present in the next reading group if interesting   🧩 What you’ll be working on • Invent next-generation, efficient generative world-models (diffusion, transformer or hybrid) that deliver real-time roll-outs and controllable scene editing. • Architect interactive world models where agents (or humans) can step the model, enabling reinforcement learning, planning and safety evaluation loops. • Optimise end-to-end performance – from latent compression to context pruning – your aim is to reduce inference latency by orders of magnitude. • Define robust metrics for long-horizon coherence, physics fidelity and planner integration; run ablations and scaling studies to understand trade-offs. • Ship impact: integrate your models into closed-loop training and evaluation, and measure the sim-to-real gap against on-road driving-model results. • Mentor and influence: guide junior researchers, shape technical road-maps, publish at top venues and represent Wayve in the community. • Challenge assumptions and drive innovation: propose bold ideas, conduct ablation studies, and question conventional approaches to training and evaluation.   🙌 You should apply if • 4+ years of experience in ML research/engineering with a focus on generative video, world models. • Deep knowledge in diffusion & latent-video models; track record of improving sampling efficiency or model throughput. • Experience working with high-dimensional temporal or spatial-temporal data (e.g., video, multi-sensor fusion). • Strong Python and PyTorch engineering fundamentals, and e

View the official posting →