DevOps engineer for multimodal AI

Building the data layer for physical and multimodal AI requires moving and processing petabyte-scale video, LiDAR, and sensor streams with zero room for infrastructure failure. This company has raised over $100M from leading global VCs and grown past 180 people in London, serving hundreds of the world's most advanced AI and robotics teams. The engineering culture is dense with top-tier researchers, quants, and systems engineers building mission-critical platform primitives from first principles.

• Architect, scale, and maintain multi-cloud Kubernetes clusters, networking, and high-throughput storage across GCP and AWS, treating infrastructure-as-code via Terraform or Pulumi as the absolute default.
• Own and modernise end-to-end CI/CD pipelines, removing developer bottlenecks and enabling squads to ship reliably multiple times a day.
• Define SLIs and SLOs for critical services handling petabyte-scale multimodal datasets, leading blameless postmortems to continuously improve system resilience.
• Enforce platform observability standards, ensuring every service is fully instrumented with distributed tracing, structured logging, metrics, and actionable alerting before it hits production.
• Partner directly with product and systems engineers to profile bottlenecks, harden service architectures, and tune system performance under peak load.

• 4+ years of hands-on platform engineering, DevOps, or SRE experience maintaining mission-critical workloads in production.
• Deep operational experience with Kubernetes, container orchestration, and multi-cloud infrastructure across AWS or GCP.
• Production mastery of infrastructure-as-code using Terraform or Pulumi, alongside a track record of building automated CI/CD pipelines at scale.
• Strong fundamentals in distributed systems observability (metrics, logs, tracing, alerting), Linux operating systems, networking, and databases.
• Based in or able to relocate to London, working on-site four days a week.

• Условий в вакансии нет