ML Infrastructure Engineer: Scalable Training and Deployment

Epsilon; San Francisco, Northern (CA, KY); Hybrid; USD 180,000 - 280,000; Full time

Job summary

Epsilon Health in San Francisco is seeking an ML infrastructure engineer to design core systems for scalable training of large models in medical imaging. You will enable researchers to run experiments efficiently, focusing on science rather than bottlenecks.

In this role you own distributed training, data loading, inference and deployment pipelines, and the RL training stack; collaborate with research and backend teams to ship production-grade solutions.

Qualifications

• 6+ years designing, building, and operating large-scale distributed systems in production.
• 2+ years building ML infrastructure or systems in production.
• Strong Python skills and expertise in PyTorch or JAX.
• Experience and familiarity with the compute, tooling, and workflow needs of large-scale ML research.
• Experience building infrastructure or platforms specifically for research or ML workflows.
• Deep experience building and operating Kubernetes and cloud infrastructure at scale.
• Experience with distributed training at scale (FSDP, DeepSpeed, or Megatron-style parallelism).
• Prior experience as a technical lead or mentor for other engineers.

Responsibilities

• Partner directly with researchers to understand workflows and design needs.
• Build a distributed training infrastructure for foundation models on large-scale medical imaging.
• Build high-throughput data loading and preprocessing to keep GPUs saturated.
• Prototype ideas and translate them into production-ready code with end-to-end delivery.
• Contribute to production serving and deployment pipelines alongside the backend team.
• Build the reinforcement learning training stack for online, multi-reward RL at scale.

Skills

Distributed systems

ML infrastructure

Python

PyTorch

JAX

Kubernetes

Cloud infrastructure

Large-scale training

Leadership

Research workflows

Tools

DeepSpeed

Megatron

TensorRT

Triton

vLLM

Job description

Epsilon Health in San Francisco is seeking an ML infrastructure engineer to design core systems for scalable training of large models in medical imaging. You will enable researchers to run experiments efficiently, focusing on science rather than bottlenecks.

In this role you own distributed training, data loading, inference and deployment pipelines, and the RL training stack; collaborate with research and backend teams to ship production-grade solutions.