Senior Network Engineer - Data Center Networking

About the Role

The Data Center Network Engineering (DNE) pod within the Physical Networking team owns the design, deployment, and operation of Roblox's large-scale data center network. This includes server-layer connectivity, intra-DC and inter-DC connectivity. The team builds High-Performance Networking (HPN) fabrics, deploys 400G/800G networking, and scales GPU clusters supporting AI training and inference.

Responsibilities

• Drive the design, deployment, and operations of Roblox's large-scale data center network, including networking for GPU clusters supporting AI training and inference workloads.
• Design, qualify, and validate new technologies and hardware for scalable solutions.
• Coordinate data center builds and projects with multiple teams, including weekly cross-team calls.
• Work with Network Reliability Systems, Data Center Ops, Capacity Planning, Supply Chain, and Compute/ML teams to address network challenges early.
• Contribute to automation and telemetry to improve efficiency, reliability, and visibility.
• Participate in an on-call rotation to support global network infrastructure 24x7.

Qualifications

Minimum Requirements

• 6+ years of expertise designing and deploying large-scale data center fabrics (CLOS).
• Experience with GPU clusters or AI/ML networking.
• Proficiency deploying and managing 400G and 800G hardware.
• Skilled in configuring and troubleshooting large-scale BGP deployments.
• Strong knowledge of network protocols (TCP/IP, UDP, DHCP, DNS), with hands-on experience in IPv4 and IPv6.

Preferred Qualifications

• Familiarity with QoS, IS-IS, MPLS, RSVP-TE, DWDM, commercial NAT, open-source hardware and software.