Network Systems Engineer

About the company

Efference provides high-performance perception and compute infrastructure for training and deploying robot models in the cloud. Our vision is to make robotic deployments scalable with shared compute.

We approach the problem by owning the entire path from the image sensor to the datacenter GPU. We shrink the load at the sensor, fix the uplink so robot traffic is fast and reliable, and clock synchronize the system so observations arrive time-aligned, ready for inference.

About the role

You'll own the multi-path network between our access points and the compute that serves robot models. You'll make sensor uploads and returning inference results fast and predictable, working with wireless, robotics, and ML infrastructure engineers to control queueing and recover from failures.

What you’ll do

• Design and implement multi-path transport across mesh backhaul, site uplinks, and datacenter hosts, with explicit budgets for latency, jitter, loss, and observation age.
• Develop traffic classification, pacing, congestion control, and backpressure that protect live inference and control traffic when recordings or other workloads compete for bandwidth.
• Tune Linux network paths and NIC behavior, profiling socket buffers, packet processing, CPU affinity, copies, and application-level queueing.
• Build connection establishment, authentication, reconnect, and failover behavior that preserves sessions through handoffs, changing routes, and customer firewalls.
• Implement cross-host timing and tracing with embedded engineers, separating clock uncertainty from actual transport delay.
• Test multi-path routing, mesh failures, beamforming changes, and handoffs with wireless engineers; measure loss, reordering, and latency under concurrent robot traffic.

What we are looking for

• Experience building production networking software or distributed systems where latency and availability matter.
• Deep understanding of TCP/IP, UDP, routing, queueing, congestion, and how transport choices affect application behavior.
• Strong systems programming in C++, Rust, or Go, plus practical Linux networking and performance debugging skills.
• Experience measuring tail latency and diagnosing problems using packet traces, host instrumentation, and realistic load tests.
• Ability to reason across application, operating system, network equipment, and provider boundaries.

Nice to Have

• Experience with WebRTC, QUIC, real-time media, or custom transport for sensor data.
• Knowledge of PTP, hardware timestamping, eBPF/XDP, DPDK, RDMA, or accelerated NIC paths.
• Experience with robotics, edge-to-cloud infrastructure, or deploying services on customer-controlled networks.