Staff ML Infra Engineer — Hybrid, Equity

Remanence; Paris; Hybrid; EUR 110,000 - 150,000; Full time

Benefits offered by this job

Visa sponsorship

Relocation support

Flexible hybrid setup

Job summary

Remanence is building the next era of enterprise AI with scalable infrastructure for training and inference. You will own the compute and execution platform, enabling efficient use of GPU resources, reproducible experiments, and robust operation across large-scale ML workloads.

The role focuses on building and operating GPU clusters, sandboxed environments, and reliable data pipelines, with a strong emphasis on observability and performance improvements for researchers.

Qualifications

• Strong systems fundamentals and ability to reason about concurrency, resource contention and failure recovery.
• Experience with distributed ML workloads and scalable infrastructure.
• Familiarity with Linux-based environments, Docker and Kubernetes.
• Ability to build and operate data and artifact pipelines for ML experiments.

Responsibilities

• Build and operate GPU clusters, job scheduling, networking, and storage for distributed ML workloads.
• Develop execution platform for large numbers of concurrent environments with sandbox isolation and recovery.
• Build data pipelines for datasets, trajectories, model weights, and checkpoints.
• Own platform observability and reliability through capacity allocation and latency improvements.
• Provide researchers with reproducible environments and simple tools to launch and debug experiments.

Skills

Systems fundamentals

Concurrency

Resource management

Tools

Docker

Kubernetes

Slurm

Terraform

Python

Go

Job description

Remanence is building the next era of enterprise AI with scalable infrastructure for training and inference. You will own the compute and execution platform, enabling efficient use of GPU resources, reproducible experiments, and robust operation across large-scale ML workloads.

The role focuses on building and operating GPU clusters, sandboxed environments, and reliable data pipelines, with a strong emphasis on observability and performance improvements for researchers.