ML Infrastructure Engineer – GPU Compute Platform

Remanence; Paris (TX); Hybrid; USD 124,000 - 186,000; Full time

Benefits offered by this job

Visa sponsorship

Relocation support

Hybrid work setup

Equity

Job summary

Remanence in Europe seeks a Member of Technical Staff - Infrastructure to own the compute and execution platform for training, inference and evaluation. You will ensure GPU resources are productive and research workflows are easy to operate and debug.

You will build and run GPU clusters, scheduling, networking and storage for distributed ML workloads; develop the execution platform for many concurrent environments; create data and artifact pipelines and improve observability, latency and

Qualifications

• Strong systems fundamentals and careful reasoning about concurrency, resources and failure recovery.
• Ability to make complex infrastructure understandable and dependable for users.
• Prior ML infrastructure experience is valuable.

Responsibilities

• Build and operate GPU clusters, job scheduling, networking and storage for distributed ML workloads.
• Develop the execution platform for large numbers of concurrent environments, including sandbox isolation, resource limits, retries, and state recovery.
• Build reliable data and artifact pipelines for datasets, trajectories, model weights, and checkpoints.
• Own platform observability and recovery; improve capacity allocation, startup latency, and reliability through measurable changes.
• Provide researchers with reproducible environments and simple tools to launch, inspect, and debug experiments.

Skills

Systems fundamentals

Concurrency

Distributed ML

Observability

Tools

Linux

Docker

Kubernetes

Slurm

Terraform

Python

Go

Ray

PostgreSQL

Prometheus

Grafana

OpenTelemetry

NVIDIA DCGM

Job description

Remanence in Europe seeks a Member of Technical Staff - Infrastructure to own the compute and execution platform for training, inference and evaluation. You will ensure GPU resources are productive and research workflows are easy to operate and debug.

You will build and run GPU clusters, scheduling, networking and storage for distributed ML workloads; develop the execution platform for many concurrent environments; create data and artifact pipelines and improve observability, latency and