Software Engineer (Scheduler), London
Core
Design, operate, and extend ML workload scheduling systems to handle dynamic batch inference and long-running training workloads on Kubernetes.
Role type
Senior IC infrastructure engineer (ML scheduling)
Builds
Custom Kubernetes controllers, operators, and CRDs for cluster orchestration
Domain
Cloud infrastructure + Machine Learning
Deliverable
infrastructure
Required skills
Kubernetes (custom controllers, operators, CRDs), Python or Go, distributed systems, batch queuing frameworks (Kueue, Volcano, Slurm), cloud infrastructure (compute provisioning, networking, IaC), DevOps/SRE practices
Preferred skills
Kueue operation, ML infrastructure nuances (Nvidia GPUs, Google TPUs, gang scheduling), High-Performance Computing, GCP compute infrastructure
Technologies
Kubernetes, Python, Go, Kueue, Volcano, Slurm, GCP
Responsibilities
Design and operate ML workload scheduling systems; optimize compute utilization via gang scheduling and topology-aware placement; collaborate with ML researchers to troubleshoot bottlenecks; build robust operational processes and SLOs; develop custom Kubernetes controllers and operators.
