Platform Software Engineer – AI/ML & GPU Infrastructure

Role details

The application window is expected to close on: 12/27/2026 Meet the Team You'll join the Platform Engineering team building and operating Cisco's AI Compute Platform — the GPU infrastructure that powers model training, inference, and LLM workloads at scale. The Platform Engineering team builds and runs the unified control plane and Kubernetes platform that serve as the foundation of this infrastructure, owns how GPU capacity is scheduled and allocated across shared and dedicated pools, and partners closely with infrastructure, hardware, and engineering teams to turn raw GPU capacity into a reliable, self-service platform for every team that needs it. Your Impact As a Platform Software Engineer on the Platform Engineering team, you will design, build, and operate the control-plane and scheduling systems that make GPU infrastructure usable at scale. You will work at the intersection of distributed control planes, Kubernetes, GPU acceleration, and workload orchestration. In this role, you will build and operate distributed systems, control-plane services, and APIs that manage multi-cluster environments, enforce tenant GPU quotas, and monitor complex training and inference workloads. Your work will blend systems software engineering, API development, and operations to maximize reliability and throughput across our AI compute fleet. Responsibilites: Design and develop platform control-plane services, APIs, and controllers to manage multi-cluster lifecycle, tenant project onboarding, and GPU quota enforcement across shared and dedicated pools. Implement GPU capacity scheduling and allocation — including fractional GPU allocation, multi-tenant isolation, and workload-aware scheduling logic that accounts for training (checkpointing, preemption tolerance) vs. inference (latency sensitivity, multi-node placement) differences. Build automated capacity-intake, demand-classification, and request-routing tooling that routes workloads to the right cluster and compute pool. Build automated remediation tooling that detects node and workload failures, triggers automated remediation, and helps optimize system-wide performance for training and inference workloads. Design, build, and maintain scalable, reliable Kubernetes-based platforms (OpenShift or upstream Kubernetes) as the foundation of the AI Compute Platform. Develop platform services and automation using Kubernetes Operators, Controllers, CRDs, Go, Python, and modern software engineering practices. Develop platform automation, internal tools, and dashboards to improve platform visibility, efficiency, and reliability. Implement Infrastructure as Code, GitOps, and CI/CD automation for the platform; participate in on-call rotation. Minimum Qualifications: Bachelor's degree in Computer Science, Computer Engineering, or equivalent practical experience. 5+ years of software engineering experience building, operating, or extending Kubernetes-based platforms or distributed systems. Strong software development proficiency in Go or Python , with hands-on experience building platform control-plane services, custom APIs, Kubernetes Controllers, Operators, or CRDs. Hands-on experience building, configuring, or operating batch and accelerator schedulers (e.g., Kueue, Volcano, Slurm , or custom scheduler plugins) for GPU or AI/ML workloads. Practical experience with GPU scheduling, multi-tenant GPU infrastructure, or managing workloads on accelerated compute platforms. Hands-on experience with Infrastructure as Code (e.g., Terraform), GitOps, CI/CD deployment pipelines, and cloud-native observability tools (e.g., Splunk, Prometheus, Grafana). Preferred Qualifications: Familiarity with multi-cluster control plane architectures, cluster registration/federation, and large-scale tenant isolation models. Familiarity with modern data center GPU architectures (e.g., NVIDIA H100, H200, B200/B300, or AMD Instinct accelerators) and their scheduling and placement considerations. Familiarity with topology-aware placement, collective-communication libraries (e.g., NCCL), and gang scheduling for multi-node training clusters. Experience supporting AI/ML infrastructure, including model serving pipelines, distributed training, workload orchestration, or MLOps platforms. Experience building automated capacity-intake, demand-classification, or request-routing tooling (rules-based or AI-assisted) for infrastructure capacity management. Experience developing engineering-focused developer tools, CLI utilities, or dashboards that simplify complex platform operations. Exposure to Kubernetes control plane internals (API server, etcd, controller-manager, scheduler framework). Experience participating in on-call rotations, incident response, and postmortem/runbook processes for production infrastructure. Why Cisco? At Cisco, we’re revolutionizing how data and infrastructure connect and protect organizations in the AI era – and beyond. We’ve been innovating fearlessly for 40 years to create solutions that power how humans and technology work together across the physical and digital worlds. These solutions provide customers with unparalleled security, visibility, and insights across the entire digital footprint. Fueled by the depth and breadth of our technology, we experiment and create meaningful solutions. Add to that our worldwide network of doers and experts, and you’ll see that the opportunities to grow and build are limitless. We work as a team, collaborating with empathy to make really big things happen on a global scale. Because our solutions are everywhere, our impact is everywhere. We are Cisco, and our power starts with you. Message to applicants applying to work in the U.S. and/or Canada: The starting salary range posted for this position is $155,000.00 to $222,600.00 and reflects the projected salary range for new hires in this position in U.S. and/or Canada locations, not including incentive compensation*, equity, or benefits. Individual pay is determined by the candidate's hiri