Network and Compute Engineer, AI Infrastructure

Описание

This role involves designing, deploying, validating, and operating high-performance infrastructure for enterprise artificial intelligence and accelerated-computing workloads.

Responsibilities

• Deploy, configure, validate, and operate accelerated-compute clusters
• Perform server and cluster bring-up, including operating-system imaging, boot validation, firmware alignment, hardware discovery, and post-deployment health checks
• Support NVIDIA MGX-based infrastructure and understand the relationship between hardware architecture, firmware, operating systems, drivers, and the platform software stack
• Install and administer Linux-based compute environments
• Support Kubernetes installation, configuration, upgrades, and lifecycle operations
• Deploy and administer Slurm-based workload environments
• Assist with NVIDIA GB200 firmware updates, software-stack validation, sanity checks, and platform readiness testing
• Support NVLink configuration and partitioning, where applicable
• Troubleshoot GPU and system errors, including XID events, degraded nodes, boot failures, firmware inconsistencies, and workload-performance issues
• Develop standardized, repeatable build and validation procedures for compute nodes and clusters
• Design and support resilient network connectivity for accelerated-compute and AI cluster environments
• Configure and troubleshoot high-speed Ethernet and, where applicable, InfiniBand fabrics
• Support leaf-spine network architectures, network segmentation, routing, switching, addressing, name resolution, and time synchronization
• Diagnose packet loss, latency, congestion, interface errors, cabling faults, routing issues, and workload connectivity failures
• Validate host-to-host, node-to-storage, management-plane, and workload-plane connectivity
• Support network configuration required for Kubernetes, cluster provisioning, telemetry, and distributed workloads
• Work with security teams to implement network isolation, firewall rules, access controls, and secure administrative paths
• Maintain accurate network diagrams, port maps, addressing plans, and configuration records
• Install, configure, and operate NVIDIA Base Command Manager
• Support rack import, rack view, base view, node provisioning, and cluster lifecycle activities
• Work with bare-metal management systems for server discovery, power operations, hardware health, firmware administration, and remote troubleshooting
• Integrate cluster platforms with enterprise directory services
• Support NVIDIA Mission Control processes and related AI Factory operational practices
• Assist with the deployment and operation of Run:ai or equivalent workload-orchestration capabilities
• Automate repetitive provisioning, validation, configuration, and reporting activities
• Implement infrastructure monitoring for compute nodes, GPUs, network devices, interfaces, firmware, thermals, power, storage dependencies, and cluster services
• Establish actionable dashboards, telemetry, alerts, and operational health checks
• Lead technical diagnosis across hardware, firmware, Linux, Kubernetes, scheduler, network, and application layers
• Perform root-cause analysis and document corrective and preventive actions
• Develop operational runbooks for failure scenarios such as node loss, degraded network links, GPU errors, provisioning failures, and cluster-service disruption
• Participate in incident response, problem management, capacity planning, lifecycle planning, and controlled change implementation
• Identify automation and standardization opportunities that reduce manual effort, deployment defects, and recovery time
• Work with infrastructure, cloud, security, storage, application, vendor, and service-management teams
• Apply least-privilege access to management, compute, network, and orchestration platforms
• Integrate administrative access with approved enterprise identity and directory services
• Secure management interfaces, service accounts, automation credentials, application programming interfaces, and remote-access pathways
• Maintain separation between management, workload, storage, and out-of-band management networks
• Support firmware, operating-system, container, Kubernetes, and network vulnerability remediation
• Ensure administrative actions and infrastructure changes are traceable through logging and change records
• Protect sensitive cluster configuration, credentials, certificates, logs, and telemetry
• Follow secure configuration standards, change controls, and applicable customer or regulatory requirements

Requirements

• Strong administration and troubleshooting experience with Linux
• Strong knowledge of data-center networking, including Ethernet, Transmission Control Protocol/Internet Protocol, routing, switching, virtual local area networks, link aggregation, Domain Name System, Dynamic Host Configuration Protocol, and Network Time Protocol
• Hands-on experience installing or operating Kubernetes clusters
• Experience deploying and supporting physical compute infrastructure
• Understanding of server boot processes, operating-system imaging, firmware, device drivers, hardware management, and remote management controllers
• Experience diagnosing issues across server, operating-system, network, and application layers
• Working knowledge of Python, Bash, Ansible, or equivalent infrastructure-automation technologies
• Experience with monitoring, logging, metrics, dashboards, alerting, and telemetry
• Ability to create clear architecture diagrams, build documents, operating procedures, and troubleshooting runbooks
• Strong analytical skills and the ability to work effectively during complex infrastructure incidents
• Bachelor
• Recommended 7 or more years of experience in infrastructure, compute, network, data-center, cloud-platform, or high-performance-computing engineering
• Recommended 3 or more years of hands-on experience with Linux-based cluster platforms, Kubernetes, accelerated computing, or comparable large-scale infrastructure
• Demonstrated experience owning complex technical problems from initial diagnosis through resolution and prevention

Nice to have

• Experience with NVIDIA AI infrastructure, NVIDIA Mission Control, or NVIDIA Base Command Manager
• Experience with NVIDIA MGX or GB200-based systems
• Experience with Slurm and distributed-compute scheduling
• Experience with Run:ai
• Experience with InfiniBand, Remote Direct Memory Access, or Remote Direct Memory Access over Converged Ethernet
• Knowledge of NVLink and multi-GPU communication
• Experience interpreting NVIDIA XID errors and GPU telemetry
• Experience with bare-metal provisioning and Redfish-compatible management interfaces
• Experience with Prometheus, Grafana, OpenTelemetry, or equivalent observability platforms
• Exposure to container runtimes, container registries, Kubernetes networking, and graphics-processing-unit operators
• Experience supporting production AI, machine-learning, high-performance-computing, or large-scale data platforms
• NVIDIA Certified Associate or NVIDIA Certified Professional credentials relevant to AI infrastructure
• Red Hat Certified System Administrator or Red Hat Certified Engineer
• Certified Kubernetes Administrator
• Cisco Certified Network Professional, Data Center or Enterprise
• NVIDIA networking or InfiniBand training
• Linux Foundation certifications
• Security-focused credentials appropriate to the operating environment

We offer

• Client is not willing to sponsor any visa
• Contract W2
• Contract Corp To Corp
• 12 Months
• Travel Required
• $60 - $65/hr

Hiring process

• Log in or create an account to see job match score