Staff ML Engineer, Foundation Models

Woven by Toyota is enabling Toyota’s once-in-a-century transformation into a mobility company. Inspired by a legacy of innovating for the benefit of others, our mission is to challenge the current state of mobility through human-centric innovation — expanding what “mobility” means and how it serves society. Our work centers on four pillars: AD/ADAS, our autonomous driving and advanced driver assist technologies; Arene, our software development platform for software-defined vehicles; Woven City, a test course for mobility; and Cloud & AI, the digital infrastructure powering our collaborative foundation. Business-critical functions empower these teams to execute, and together, we’re working toward one bold goal: a world with zero accidents and enhanced well-being for all. THE TEAM At Woven by Toyota, we work on a diverse set of problems ranging from solving optimization problems in 3D geometric computer vision to designing and deploying novel ML architectures for perception, prediction, and motion planning for Toyota customers. We are looking for doers and creative problem solvers to join us in improving mobility for everyone with self-driving technology. You will be interacting on a daily basis with other software and hardware engineers and researchers to tackle some of the most challenging problems in AI, robotics, and computer vision. WHO ARE WE LOOKING FOR? The AD/ADAS team is seeking a seasoned ML engineer to support the development of foundation models for our autonomy stack. You will be responsible for curating the right multi-modal data, identifying the most suitable model architectures, steering towards the best training paradigms, optimizing the infrastructure for the highest utilization, and ultimately focusing on efficient deployment for onboard and offboard use cases. Fortunately, you will be doing this with a world-class team of exceptional engineers. We recognize the unique capabilities each team member brings and encourage applicants to reach out even if they do not match all of the characteristics described below. RESPONSIBILITIES

Lead the roadmap, research, development, and deployment of foundation models / VLAs / VLMs / world models for our autonomy stack across a variety of applications

Discover new ways to leverage heterogeneous in-house data at Toyota scale

Architect and implement efficient, large-scale distributed ML pipelines for training and fine-tuning, as well as pre/post-training of large multi-modal, multi-task models 

Design experiments and benchmarks to evaluate foundation models for autonomy scenarios

Optimize models for  speed, accuracy, and robustness under diverse environmental conditions

Translate cutting-edge research into useful production-ready solutions

Enable and help other engineers on the team to be more effective through coaching and leading by example when it comes to writing high-quality code, providing high-quality code and design

Work in a high-velocity environment and employ agile development practices

Team player and “get things done” mentality

Closely collaborate with internal and external teams  to develop end-to-end solutions

MINIMUM QUALIFICATIONS

M.Sc. or higher degree in CS/CE/EE, or equivalent industry experience.

8+ years of experience with popular ML frameworks such as PyTorch or similar (e.g. JAX)

6+ years of experience in efficient ML training & deployment workflows: data sampling and sharding, pre-processing, distributed training and parallelism approaches, model training, evaluation & ablation, inference optimization

Prior proven experience in training and deploying state-of-the-art large foundation models leveraging large-scale data 

Strong programming skills in Python and C++

Passionate about advancing self driving technology for the mass market

Strong communication skills and ability to communicate concepts clearly and precisely

Experience in production-level software development best practices.

NICE TO HAVES

Hands-on experience with building a perception stack for autonomous systems

Experience in runtime optimization for Linux and UNIX-like real-time operating systems

Publications at major conferences such as CVPR, ICCV, ECCV, ICRA, NeurIPS etc. 

Experience with embedded CUDA development and related tooling

Experience with automotive-grade edge-compute platforms and safety-critical software