Job Description
This job was selected by the SpringCube team to help AI, Data and Cloud Engineers discover relevant San Francisco Bay Area employers. Sign up to view the full employer details and apply directly with the hiring company.
Company Overview
The company develops general-purpose artificial intelligence systems designed to operate in the physical world. Its work brings together machine learning research, robotics, software engineering, and advanced hardware to build foundation models and learning algorithms for robots and other physically actuated devices.
Its engineering teams focus on developing scalable AI models and infrastructure that enable intelligent machines to perform increasingly complex real-world tasks.
Responsibilities
- Training and Inference Infrastructure
- Design, implement, and maintain infrastructure for large-scale machine learning training.
- Develop and manage systems for job scheduling, workload orchestration, checkpointing, metrics collection, and logging.
- Ensure training infrastructure remains reliable, efficient, and scalable.
- Distributed Training
- Collaborate with researchers to scale JAX-based training workloads across GPU and TPU clusters.
- Improve distributed training workflows to minimize operational complexity and support efficient experimentation.
- Work with multi-host computing environments and distributed training configurations.
- Performance Optimization
- Profile and optimize memory consumption, device utilization, training throughput, and distributed synchronization.
- Identify performance bottlenecks across the machine learning training stack.
- Improve resource efficiency and overall training performance.
- Experimentation and Developer Tools
- Build reusable abstractions for launching, monitoring, debugging, and reproducing machine learning experiments.
- Improve workflows that enable researchers and engineers to iterate quickly and reliably.
- Develop tools that simplify experiment management and troubleshooting.
- Research Collaboration
- Work closely with researchers and model engineers to translate research requirements into infrastructure capabilities.
- Establish and promote best practices for large-scale model training.
- Support the transition of research experiments into reliable, production-grade training workflows.
- Core Training Code Development
- Improve and extend JAX-based model and training code.
- Support the implementation of new model architectures, data modalities, and evaluation metrics.
- Contribute to the development of efficient and maintainable machine learning systems.
Qualifications
- Strong software engineering fundamentals.
- Experience building machine learning training infrastructure, internal developer platforms, or related systems.
- Hands-on experience with large-scale model training using JAX, preferably, or PyTorch.
- Familiarity with distributed training, multi-host computing environments, data loaders, and model evaluation pipelines.
- Experience managing machine learning training workloads on cloud or cluster platforms such as SLURM, Kubernetes, Google Cloud TPU/GKE, or AWS.
- Ability to identify, debug, and optimize performance bottlenecks across the machine learning training stack.
- Strong cross-functional communication skills and a high level of ownership.
- Ability to collaborate effectively with research, data, platform, and model engineering teams.
Preferred Qualifications
- Deep background in machine learning systems, including training compilers, runtime optimization, or custom kernel development.
- Experience working close to hardware through GPU or TPU performance tuning.
- Background in robotics, multimodal machine learning, or large-scale foundation models.
- Experience designing infrastructure abstractions that balance researcher flexibility with system reliability and operational efficiency.
Disclaimer
SpringCube curates tech job listings from various company websites to support tech professionals globally.
- No Endorsement: Job ads on SpringCube do not imply endorsement of their authenticity or quality.
- No Client Relationship: This company is not a client of SpringCube unless stated.
- To Apply: Click the “Apply” button to be redirected to the hiring company’s application page for this job.
- No Liability: SpringCube is not liable for inaccuracies.