ML Infrastructure Engineer

We’re supporting a client with the recruitment requirement outlined below. If your recruiting expertise and networks align with the role, please 'apply' to express your interest.

Once approved to work on the opportunity, you’ll receive the full brief, including the client name, role details, and commercial terms.

Role Details

Company: US-based robotics and AI startup

Location: Redwood City, California

Employment Type: Full-time

Work Arrangement: On-site, 5 days a week

Salary: $220,000–$320,000, with flexibility for exceptional candidates

Experience: 5+ years

Visa Sponsorship: Available

About the Company

An advanced robotics company developing general-purpose robots powered by proprietary AI models. Its technology is already deployed in commercial environments, including hospitality and restaurant industries.

The company is backed by leading technology investors and brings together experienced engineers and researchers from major AI, technology, and autonomous systems companies.

About the Role

The company is hiring an ML Infrastructure Engineer to build and optimise the infrastructure supporting large-scale machine learning training.

This role focuses on distributed training systems, GPU clusters, and high-throughput data pipelines for multimodal AI models. The ideal candidate combines strong software engineering skills with hands-on experience operating large-scale ML training infrastructure.

Responsibilities

• Build and optimise distributed ML training infrastructure across GPU clusters.

• Improve training performance using PyTorch Distributed, FSDP, CUDA, and NCCL.

• Manage cluster scheduling, orchestration, and failure recovery.

• Profile GPUs and troubleshoot multi-node performance bottlenecks.

• Develop high-throughput pipelines for multimodal sensor data.

• Improve infrastructure scalability, reliability, and training efficiency.

• Collaborate with research and engineering teams to support advanced AI model development.

Requirements

• 5+ years of infrastructure or software engineering experience.

• Proven ownership of large-scale distributed ML training systems.

• Hands-on experience with PyTorch Distributed, FSDP, GPU sharding, CUDA, and NCCL.

• Experience operating research compute clusters using SLURM or Kubernetes.

• Experience working with real-world sensor or multimodal data, such as video, audio, and robotics data.

• Strong programming and software engineering fundamentals.

• CS or related technical degree from a strong program

• Willingness to work on-site in Redwood City five days a week.

Interview Process

• Candidate submission and hiring manager review.

• Recruiter screening.

• System design interview.

• Practical coding interview.

• Second coding interview.

• Behavioral and cultural fit interview.

• Offer and hiring.