Staff Machine Learning Software Engineer

Wayve•Sunnyvale, CA

6d•$311,000 - $419,000•Hybrid

About The Position

Wayve is seeking a founding member for its Multi-Embodiment Generalist Agent (MEGA) team within Wayve Science. The MEGA team is focused on building foundation models for general-purpose robots, utilizing large-scale video, language, and robot-interaction data to generalize across various tasks and embodiments. The role involves designing and building the software systems essential for MEGA's machine learning research, including scalable ML pipelines, data and training infrastructure, and shared tools. The position is hands-on with large-scale video and robotics datasets, modern ML frameworks, and increasingly large models, emphasizing the development of reliable and maintainable systems. A key aspect of the role is designing and maintaining robust workflows and reusable infrastructure to support multiple ML projects, with a strong focus on infrastructure, software systems, repo health, testing, and maintainability. The successful candidate will be a core member of the MEGA team, owning critical parts of the software and ML infrastructure, working directly on systems for building, training, evaluating, and scaling models, and influencing the team's ML stack development as the program grows.

Requirements

Strong software engineering skills and experience building high-quality, maintainable software.
Experience building and maintaining machine learning pipelines or infrastructure, such as data ingestion, training, evaluation, or experiment workflows.
Experience designing software systems and abstractions that are reliable, reusable, and able to evolve as requirements change.
Hands-on experience with modern machine learning frameworks and a good understanding of how ML training and experimentation workflows operate.
Strong debugging skills and the ability to investigate problems across complex ML systems.
Experience with software testing, code quality, and engineering practices for maintaining a healthy shared codebase.
Experience working with large datasets, large models, or other computationally demanding ML workloads.
Ability to collaborate closely with researchers and engineers and translate research requirements into practical software systems.

Nice To Haves

Experience with distributed training, multi-node systems, or large-scale data processing.
Experience supporting foundation-model training or other large-scale ML research.
Experience with multimodal models, video models, vision-language models, or related ML systems.
Experience working with robotics, embodied AI, simulation, or robot-interaction data.
Experience building infrastructure in a fast-moving applied research environment where requirements evolve rapidly.
Experience improving the performance, reliability, or developer experience of ML training and experimentation systems.

Responsibilities

Design, build, and maintain scalable ML pipelines for data ingestion, model training, evaluation, and related research workflows.
Build software systems that allow ML workloads to scale to larger datasets, larger models, and more experiments.
Develop clean and reusable interfaces between data, models, training, evaluation, and downstream robotics workflows.
Own and improve the health of the MEGA codebase, including software architecture, testing, reliability, maintainability, and engineering standards.
Identify and resolve performance, reliability, and usability bottlenecks across ML workflows.
Build and maintain infrastructure that supports multiple researchers and ML projects without unnecessarily slowing down iteration.
Work closely with researchers to understand the requirements of new models and experiments, and translate those requirements into practical software solutions.
Build and use distributed training and data-processing pipelines for large models and large multimodal datasets.