Platform Engineer

Cubiq Recruitment; Greater London; Hybride; GBP 90 000 - 130 000; Plein temps

Avantages offerts par ce poste

Bonus

Equity

Résumé du poste

Cubiq Recruitment is representing an early-stage AI research lab in London seeking a Platform/DevOps Engineer to design and build a scalable platform. You will own the Kubernetes estate, manage GPU scheduling, and develop infrastructure across GCP and other cloud providers. Strong Python skills and Rust experience are essential.

You’ll improve developer tooling, logging/monitoring, and create dashboards used by researchers to spot issues quickly. Hybrid work model offered.

Qualifications

• Kubernetes in production at scale
• Built and owned monitoring and logging for complex systems
• Strong cloud infra experience (GCP preferred)
• Production-quality Python code
• Rust experience
• Experience with Cody tooling
• Exposure to ML workloads and GPU scheduling
• Improved engineers’ tooling and local environments
• Platform/DevOps or infrastructure background in big tech or AI labs

Responsabilités

• Make core design decisions while the platform is being built
• Run and grow the Kubernetes estate and internal tooling
• Manage GPU workloads across a large cluster with schedulers like Kueue and KAI
• Investigate hardware failures and implement fixes to prevent recurrence
• Build and maintain infrastructure on GCP and other clouds
• Set up logging, metrics and dashboards used by engineers and researchers
• Improve developer environments and tooling to speed up research
• Write platform code primarily in Python, with Rust where fitting
• Take on roles where the research team needs core platform support

Connaissances

Kubernetes production

Monitoring & logging

Cloud infrastructure

Python

Rust

GPU scheduling

DevOps tooling

Platform experience

Research lab experience

Outils

Cody

Description du poste

Our client is an early-stage AI research lab in London working on fundamental questions in machine learning. Their research runs at very large scale, and compute and infrastructure are central to how quickly the team can test ideas and make progress. They’re now building out their platform team, and you’d be one of its first hires.

The role:

• You’ll make core design decisions while the platform is still being built, not maintain someone else’s setup
• You’ll run and grow the Kubernetes estate, including containerised services and internal tooling
• You’ll manage how GPU workloads are queued and allocated across a large cluster, using schedulers such as Kueue and KAI, so the hardware is used well
• When hardware fails or the cluster has an incident, you’ll investigate the cause and put fixes in place that stop it coming back
• You’ll build and maintain infrastructure on GCP, alongside other major cloud providers
• You’ll set up logging, metrics and dashboards that engineers and researchers actually use to spot and fix problems
• You’ll improve the internal developer setup (networking, machine provisioning, containerised dev environments) so researchers spend less time fighting their tools
• You’ll write the code behind the platform, mainly in Python, with Rust where it fits
• You’ll have room to take the role where the research team needs it most

About you:

• You’ve run Kubernetes in production at a serious level
• You’ve built and owned monitoring and logging for complex systems
• You know cloud infrastructure well; GCP is preferred, but AWS experience is also welcome
• You write solid, production-quality Python
• Rust experience
• Experience with Cody
• Some exposure to ML workloads, particularly how GPU jobs are scheduled and shared
• Experience improving engineers’ day-to-day tooling and local environments
• You’ve worked in platform, DevOps or infrastructure at a well-known big tech company or major AI lab

What can you expect?

• Market-leading compensation (bonus + equity)
• Working model: Hybrid