ML Platform Engineer

About this role

Foxglove seeks an ML Platform Engineer to design and operate production infrastructure for a robotics data platform handling petabyte-scale multimodal data. You'll own inference serving, embedding pipelines, training infrastructure, and cloud systems that enable rapid ML iteration across distributed robotics deployments.

What you'll do

• Design and operate production inference infrastructure including model serving, autoscaling, and cost optimization
• Own embedding and retrieval pipeline architecture for multimodal robotics data (images, video, point clouds, timeseries)
• Build training and evaluation infrastructure with job orchestration, experiment tracking, and dataset versioning
• Make cloud infrastructure decisions (AWS/GCP) balancing latency, throughput, reliability, and cost
• Create platform abstractions enabling product engineers to ship ML features without infrastructure management
• Evaluate and integrate third-party ML infrastructure components with clear build vs. buy frameworks

What they're looking for

• Production ML infrastructure and model serving (vLLM, Triton, TorchServe)
• Distributed systems and cloud infrastructure (AWS/GCP)
• Vector databases and retrieval systems at scale
• Model optimization and cloud cost management
• ML orchestration and experiment tracking
• System reliability and operational design
• Communication across ML and engineering teams
• Fine-tuning and domain adaptation for LLMs (bonus)

Benefits

• $300 monthly commuter/workspace budget
• Competitive equity in Series B company
• 100% medical, dental, vision, and life insurance for employees; 75% for dependents
• 401(k) matching up to 4%
• 4 weeks vacation plus holidays and winter break
• All-expenses-paid company off-sites twice yearly