SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Nearmap is seeking a Principal Cloud Platform Engineer to own the cloud infrastructure foundation that powers Nearmap AI. You will lead a team of up to three engineers, set technical direction for the infrastructure layer beneath the machine learning platform, and remain hands-on in the systems you design.
Nearmap processes aerial imagery at massive scale—thousands of EKS nodes running batch inference, GPU fleets spanning AWS and GCP, and production model endpoints serving customer products. Your role is to make this infrastructure reliable, secure, and cost-effective.
You will report to the Director of AI Systems Engineering and work closely with the Principal AI Platform Engineer. While they build the ML platform (workflow orchestration, distributed training, serving, LLMOps), you build the foundation: Kubernetes fleets across AWS and GCP, GPU-capable compute, networking, identity, infrastructure as code, CI/CD, observability, and security.
Key responsibilities include:
- Own technical direction for cloud infrastructure across the AI platform: Kubernetes fleet design on EKS and GKE, compute, networking, identity, and storage
- Design and operate GPU-capable compute (node pools, drivers, device plugins, autoscaling, multi-tenant isolation)
- Make critical infrastructure design decisions and lead technical evaluations
- Lead and mentor a team of up to three platform engineers; own technical hiring
- Write and own Terraform for the entire footprint; build self-service provisioning paths
- Build CI/CD and observability foundations: pipelines, metrics, logs, traces
- Own reliability and cost: SLOs, cloud spend across both clouds, capacity planning, incident response (on-call)
- Co-own standards and paved paths with the wider Nearmap developer platform team
- Implement cloud security: least-privilege identity, secrets management, network segmentation, supply chain security, encryption, and key management
- Enforce security controls in code and pipeline
Nearmap offers hybrid flexibility, four additional "YOU" days off annually, company-sponsored volunteering, generous parental leave, a work-from-overseas policy, discounted health insurance, and a Nearmap subscription. The Sydney CBD office includes stocked kitchens, in-office lunches on Tuesdays and Thursdays, and shower facilities.
REQUIREMENTS:
- 10+ years in software or infrastructure engineering, with at least 4 years focused on building and operating large-scale cloud platforms or distributed systems in production
- Proven experience leading technical projects and mentoring or managing a team of engineers
- Deep, production-level expertise designing, building, and operating Kubernetes at scale (EKS or GKE), including node pool and GPU compute management
- Hands-on experience building and managing cloud infrastructure on AWS and GCP with infrastructure as code (Terraform or Pulumi)
- Strong grounding in cloud networking, identity and access management, and secrets management with judgment to design for least privilege
- Production-quality Python or Go for platform tooling, and practical depth in modern software practice: Git, CI/CD, monitoring, alerting, automated testing
- Bachelor's or Master's degree in computer science, engineering, or related technical field, or equivalent practical experience
HIGHLY DESIRABLE:
- GPU infrastructure at fleet scale: NVIDIA drivers and device plugins, MIG or time-slicing, multi-cloud capacity strategy
- Multi-tenant Kubernetes: namespace isolation, resource quotas, admission policies, custom operators or Crossplane
- Infrastructure beneath ML workloads such as Argo Workflows, Ray, or Kubeflow
- GitOps and progressive delivery at scale: ArgoCD or Flux with canary and blue-green strategies
- FinOps across multiple clouds: tagging, rightsizing, commitment strategy, GPU utilization reporting
- Cloud security and compliance: SOC 2 and NIST Cyber Security Framework
- Petabyte-scale data pipelines or geospatial and computer vision workloads
