ML Networking Systems Engineer - Low-Latency, Kernel & RDMA

Anthropic; New York (NY); Hybrid; USD 280,000 - 850,000; Full time

Job summary

Anthropic is seeking a systems-level engineer specializing in network infrastructure to build and optimize software interfacing accelerators with high-speed networks.

You will work across kernel and user-space networking, dive into TCP/IP stack internals, XDP, eBPF, DPDK, and RDMA, and help design new network protocols for large-scale systems. A Bachelor’s degree is required and hybrid office presence is expected.

Qualifications

• Expertise in network protocols and networking concepts.
• Deep kernel networking: TCP/IP stack internals, XDP, eBPF, io_uring, and epoll.
• User-space networking: DPDK, RDMA, kernel bypass techniques.
• Knowledge to build abstractions like collectives and RPC for distributed systems.
• Experience diagnosing networking issues in OSI layers 2-4.

Responsibilities

• Write and maintain software interfacing accelerators with high-speed networks.
• Diagnose and optimize distributed networking software across kernel and user space.
• Design and implement network protocols and performance optimizations.
• Collaborate with ML accelerator teams to improve data movement over networks.

Skills

Networking protocols

Kernel networking

User-space networking

OS/driver programming

Rust familiarity

Education

Bachelor's degree

Tools

DPDK

RDMA

io_uring

epoll

XDP

PCIe drivers

Job description

Anthropic is seeking a systems-level engineer specializing in network infrastructure to build and optimize software interfacing accelerators with high-speed networks.

You will work across kernel and user-space networking, dive into TCP/IP stack internals, XDP, eBPF, DPDK, and RDMA, and help design new network protocols for large-scale systems. A Bachelor’s degree is required and hybrid office presence is expected.