Staff ML Infrastructure Engineer
Job Summary
As a Staff ML Infra Engineer on the Offboard Perception team, you will develop and deploy offboard machine learning solutions that provide ground-truth-quality world estimates. You will design, build, and maintain robust ML infrastructure to support autonomous vehicle development, from training and validation to deployment. This role involves cross-functional collaboration to transition research into scalable onboard ML capabilities while enhancing the overall autonomy stack.
Responsibilities
- Design and maintain ML infrastructure for perception models
- Implement CI/CD pipelines for machine learning systems
- Develop evaluation and observability frameworks
- Support dataset curation and model integration
- Mentor engineering team and enforce best practices
Required Skills
- Python (PyTorch, NumPy)
- ML pipeline development
- Kubernetes and Docker
- Distributed training and GPU infrastructure
- CI/CD and model monitoring
Job Details
What You'll Do:
- Design, build, and maintain ML infrastructure that enables rapid development, training, evaluation, and deployment of offboard perception models.
- Own the integration of models into production systems, including packaging, validation, deployment, and rollout strategies.
- Implement CI/CD pipelines for ML systems, including automated testing, model validation, performance regression checks, and deployment automation.
- Establish model evaluation and observability frameworks, including training metrics, inference performance metrics, data quality checks, and production monitoring dashboards.
- Develop infrastructure for experiment tracking and benchmarking, enabling teams to compare model architectures, datasets, hyperparameters, and training procedures in a reliable and repeatable way.
- Support efficient dataset curation and ingestion pipelines that help prioritize high-value data, accelerate iteration cycles, and improve model performance on hard-edge cases.
- Partner with ML engineers, researchers, and software teams to ensure models can be reliably integrated into larger autonomy stacks and production services at scale.
- Define and enforce best practices for ML systems engineering, including reproducibility, configuration management, artifact management, security, and operational readiness.
- Support technical collaboration through code reviews, design reviews, and mentorship, helping raise the quality and maintainability of ML infrastructure across the organization.
Your Skills & Abilities:
- Strong software engineering fundamentals, including experience building reliable, maintainable, and scalable production systems.
- Proficiency in Python, with experience using ML and scientific computing libraries such as PyTorch, NumPy, and related tooling.
- Experience building and supporting ML training and deployment pipelines, including data processing, experiment execution, model packaging, and production rollout.
- Experience deploying ML models into production environments, with understanding of end-to-end workflows such as validation, serving, monitoring, and lifecycle management.
- Familiarity with distributed training and large-scale compute infrastructure, including GPUs, cluster scheduling, and performance optimization for training workloads.
- Experience with containerization, orchestration, and automation tools such as Docker, Kubernetes, workflow schedulers, and CI/CD systems.
- Experience with model observability and operational metrics, including training metrics, inference performance, reliability monitoring, and data/model drift detection.
- Strong communication and collaboration skills, with the ability to work effectively across ML, software, data, and systems engineering teams.
- Experience in robotics, perception systems, or autonomous driving is preferred.