Staff ML Engineer, Gaia

• Own and drive work on training and improving frontier-scale video world models trained in-house. • Work on Gaia, Wayve's video world model trained on large-scale driving video that predicts future frames from past context, functioning as a simulator for generating synthetic scenarios including rare or safety-critical events. • Take on a high-impact role with the opportunity to tech-lead a key area and help shape the next version of Gaia.

• Have in-depth experience training large-scale models (language, video, or other foundation models), including ownership of training at scale. • Have a strong understanding of model architecture and the ability to contribute meaningfully to architectural and training decisions. • Have strong hands-on engineering skills with modern ML stacks such as PyTorch, including debugging and performance/reliability-minded development. • Have relevant industry experience, typically 4–5+ years; advanced degrees are valued but depth of applied experience is important.

• Have direct experience with world models, video generation, or long-horizon prediction. • Have experience improving data/training pipelines and working across infrastructure constraints including distributed training, efficiency, and reliability. • Have proven technical leadership including tech lead ownership, mentoring, and setting direction across an area.

• Lead and execute large-scale training runs for video or adjacent foundation models, from experimental design through production-grade execution. • Contribute to model architecture and training strategy using first-principles understanding rather than off-the-shelf application. • Improve world-model capabilities that enable synthetic scenario generation and downstream evaluation and training of the driving model. • Partner closely with research, applications, simulation engineering, and cloud/infrastructure teams to deliver end-to-end impact. • Provide technical leadership through mentorship, review, and setting high engineering and research standards.

Wayve is one of the most exciting AI companies in the UK right now. They are tackling one of the hardest problems in technology, teaching machines to drive, and they are doing it with an approach that the rest of the industry is now converging towards. Backed by SoftBank, Microsoft, NVIDIA, Uber, Mercedes-Benz, Nissan, and Stellantis, Wayve has raised $2.8 billion in total funding and reached a valuation of $8.6 billion. With over 1,000 employees across London, Silicon Valley, Vancouver, Leonberg, Herzliya, and Tokyo, Wayve is scaling fast while keeping its London HQ at the centre. What stands out on Glassdoor (4.4/5 from 112+ reviews) is how consistently employees praise the culture, the calibre of colleagues, and the quality of the technical work. People describe it as some of the most interesting work of their careers. Wayve also offers an on-site chef, private healthcare, competitive pay with equity, and a genuine learning environment where you work alongside world-class ML researchers and engineers. That said, some reviews flag that the pace can be intense and that working across global time zones can stretch working hours. If you are an engineer, researcher, or operator who wants to work on genuinely frontier technology with real-world impact, and you thrive in fast-paced, mission-driven environments, Wayve is a rare opportunity.

• Work in a hybrid model from Wayve's London office. • Join a well-funded company backed by leading investors, founded in 2017 and pushing the boundaries of embodied AI. • Work in an inclusive environment that values diversity and embraces new perspectives.

Founded in 2017, Wayve is the leading developer of Embodied AI technology. Its advanced AI software and foundation models enable vehicles to perceive, understand, and navigate any complex environment, enhancing the usability and safety of automated driving systems. Gaia is Wayve's video world model, trained on large-scale driving video to predict future frames from past context, functioning as a simulator that helps generate synthetic scenarios including rare or safety-critical events.