Director of Infrastructure Engineering

Responsibilities

• Own Core Infrastructure & SRE: Lead multiple engineering teams responsible for Site Reliability Engineering, networking, and storage. Establish rigorous SRE practices, driving SLA/SLO definitions, incident response, observability, and automated remediation.
• Architect HPC & Global Networking: Oversee the design, scaling, and operation of Runpod’s global network backbone, as well as ultra-low-latency HPC cluster networks. Drive the implementation and optimization of InfiniBand and RDMA over Converged Ethernet (RoCE) to support massive, multi-node GPU training workloads.
• Drive Storage Engine Innovation: Direct the architecture and performance tuning of highly scalable, distributed storage systems. Ensure our storage engines can deliver the massive IOPS and throughput required to keep high-end GPUs fed with data during deep learning tasks.
• Build a High-Output Org: Hire, mentor, and grow highly technical engineering managers and senior ICs (network architects, systems engineers, SREs). Create a culture of ownership, operational excellence, and craft in a remote-first environment.
• Translate Scale into Strategy: Partner with Program Management and Product to forecast capacity requirements, shape technical roadmaps, and convert massive scale challenges into clear technical scopes, milestones, and measurable outcomes.
• Continuously Improve Systems & Flow: Drive measurable improvements in infrastructure reliability and delivery metrics, such as deployment frequency, MTTR (Mean Time To Recovery), infrastructure as code (IaC) coverage, and system uptime.
• Architectural Stewardship: Provide architectural oversight for bare-metal provisioning, virtualization layers, network fabrics, and storage clusters, ensuring seamless scalability without becoming a bottleneck for your teams.
• Cross-Functional Partnership: Coordinate cleanly with product delivery and platform teams to ensure the infrastructure primitives they rely on are robust, well-documented, and highly available.

Requirements

• 7+ years leading software, infrastructure, SRE, or networking teams, including managing managers and multiple squads, with a proven record of scaling high-availability cloud environments.
• 8+ years building and operating large-scale distributed systems, bare-metal infrastructure, or public/private cloud platforms.
• Proven hands-on background or strong architectural understanding of ultra-low latency networking. Deep familiarity with InfiniBand and/or RoCE, spine-leaf architectures, and global WAN routing protocols (BGP).
• Experience building, operating, or tuning high-performance distributed storage systems and parallel file systems (e.g., Ceph, Lustre, Weka, NVMe-oF) capable of handling heavy AI/ML I/O loads.
• Strong foundation in reliability engineering, infrastructure-as-code (Terraform, Ansible), container orchestration (Kubernetes), and modern observability stacks.
• Experience building culture, accountability, and momentum across distributed technical teams.
• Clear written and verbal communication, strong stakeholder management, and calm, decisive leadership during high-stakes operational incidents.

Preferred Qualifications

• Direct experience architecting and operating infrastructure specifically optimized for massive GPU clusters and AI/ML workloads.
• Deep understanding of hardware architectures, GPU interconnects (NVLink), and datacenter topology.
• Track record of scaling infrastructure teams in hyper-growth startup environments.
• Open-source contributions or active recognition within the infrastructure, networking, or Kubernetes communities.

Compensation & Benefits

• Competitive base pay ranges from $225,000 - $325,000 (may be inclusive of several career levels; narrowed during interview process based on experience, qualifications, and location).
• Meaningful equity in a fast-growing company (stock options for everyone).
• Generous medical, dental & vision plans.
• Flexible PTO.
• $1,200 Home Office & Equipment Stipend.
• Remote-first with Slack as main communication tool.