What You'll Be Doing
- Build and run a multi-tenant GPU-as-a-Service platform for AI workloads
- Develop GPU orchestration and scheduling using Kubernetes and Slurm
- Manage GPU-enabled Kubernetes clusters with advanced workload scheduling
- Write Infrastructure-as-Code with Terraform and Ansible
- Deliver self-service APIs, CLI tools and portals for GPU provisioning
- Create CI/CD pipelines for platform services, backed by strong monitoring and observability
- Work with teams across AI training and inference
- Improve GPU utilisation, performance and cost efficiency
- Own reliability, security and tenant isolation, including incident response and ongoing improvement
Looking for
- 4-8+ years in platform engineering or SRE
- Deep experience with Kubernetes and cloud-native infrastructure at scale
- Background in GPU workloads, distributed systems, CI/CD and automation
- Strong communication and stakeholder skills
- Comfortable juggling priorities in a fast-moving environment
- Proactive, collaborative and solutions-focused
- Sharp focus on performance, cost, security and operational detail
Bonus Points
- GPUaaS or HPC platform experience
- Familiarity with PyTorch, TensorFlow, Ray, Kubeflow or Airflow
- Knowledge of CUDA and NCCL; cloud certifications are a plus
Originally posted on