ML Platform Engineer — Scalable GPU & Kubernetes

Socket.dev; Palo Alto (CA); On-site; USD 180,000 - 280,000; Full time

Benefits offered by this job

Healthcare coverage

Relocation support

Retirement plans

Wellness programs

Meal & transportation allowances

Job summary

Mistral is seeking an ML Platform Engineer to build and operate the platform powering large-scale training, evaluation, and batch inference. You will develop infrastructure enabling researchers and engineers to run distributed GPU workloads reliably across clusters, hardware types, and regions.

You will own critical systems across the ML lifecycle, from workload scheduling to production operations, improving reliability and developer experience in a fast-moving frontier AI environment.

Qualifications

• 4+ years of experience in ML infrastructure, distributed systems, or Kubernetes platform engineering.
• Proficient in Python or Go and productive with distributed systems.
• Strong Kubernetes knowledge: controllers, operators, CRDs, scheduling, networking, storage, resource management.
• Familiar with GPU workloads and related technologies (PyTorch, CUDA, NCCL).

Responsibilities

• Build the ML Platform: services, APIs, controllers, and tooling for training, evaluation, fine-tuning, and batch inference.
• Orchestrate GPU workloads: queueing, admission control, quotas, priorities, preemption, topology-aware placement.
• Manage compute capacity: provisioning and allocation across clusters.
• Enable multi-cluster execution: placement based on capacity and locality.
• Improve researcher experience with self-service workflows for launching and debugging distributed workloads.
• Optimize performance: GPU utilization, scheduling latency, startup time, throughput.
• Build for reliability: observability, failure recovery, capacity planning, tooling.
• Operate what you build: participate in on-call rotations and troubleshoot across infra.

Skills

Python

Go

Kubernetes

Distributed systems

GPU infra

Tools

Kueue

Karpenter

Volcano

Kyverno

Job description

Mistral is seeking an ML Platform Engineer to build and operate the platform powering large-scale training, evaluation, and batch inference. You will develop infrastructure enabling researchers and engineers to run distributed GPU workloads reliably across clusters, hardware types, and regions.

You will own critical systems across the ML lifecycle, from workload scheduling to production operations, improving reliability and developer experience in a fast-moving frontier AI environment.