Network Software Engineer

Department: Engineering

Location: San Francisco

Compensation: $250K – $300K • 0.1% – 0.5%

Employment Type: FullTime

The role

We're looking for a world-class Network Software Engineer to architect and build the networking foundation of the first AI-native cloud.

You’ll design and implement the full virtual networking stack for Blaxel’s VPC—our next-generation SDN fabric optimized for agentic AI workloads. Your work will span ultra-high-performance dataplane systems (VPP, DPDK, eBPF) and the distributed control-plane and traffic services written in Rust and Go.

You’ll also expand our globally distributed networking suite—including our Rust-based load balancer, Rust HTTP proxies, traffic routing systems, and data plane components that already process 1.5M+ requests daily. Your mission: build a VPC that surpasses traditional clouds by an order of magnitude in throughput, latency, visibility, and security enforcement.

If you’re a systems-focused engineer who loves Rust, Go, networking internals, and building cloud primitives from scratch, this role is for you.

What you'll do

You’ll collaborate closely with founders and infrastructure teams to design, build, and operate Blaxel’s end-to-end SDN stack.

• Architect and implement the high-performance dataplane for Blaxel’s VPC using VPP, DPDK, eBPF, XDP, and Rust.

• Build and extend distributed control-plane systems in Go and Rust: IPAM, L2/L3 routing, topology services, ACL engines, NAT, VPC abstractions.

• Develop Rust microservices running at extreme performance envelopes—sub-millisecond latency, multi-million flow tables, multi-Gbps throughput.

• Extend our Rust load balancer and Rust HTTP proxies to support worldwide deployment and growing request volumes.

• Operate networking workloads on bare-metal clusters with large-scale IPv6-only addressing.

• Build testing, simulation, and benchmarking frameworks for both control-plane and dataplane components.

• Instrument networking systems with deep observability: metrics, flows, packet traces, latency profiles.

• Debug and resolve complex networking issues across layers L2–L7.

• Partner with compute, storage stakeholders to design cross-stack performance and reliability guarantees.

• Continuously evolve the networking strategy, introducing new primitives optimized for AI-native traffic patterns.

Who you are

• Systems-driven engineer: You understand concurrency, memory, CPU profiles, and low-level networking behavior.

• Rust and/or Go engineer: You’re strong in Rust, Go, or both—and want to engineer high-performance distributed systems.

• Networking-fluent: You deeply understand packets, queues, routing, overlays, firewalls, and how networks behave under stress.

• Performance obsessed: You think in terms of cycles, cache locality, tail latencies, NUMA effects, and throughput curves.

• Zero-to-one builder: You thrive in environments where you invent the primitives, not glue together existing ones.

• High ownership: You want to design systems from idea → architecture → implementation → deployment → operations.

Required skills

• 3+ years in systems engineering, distributed systems, networking, or infrastructure software roles

• Strong proficiency in Rust, Go, or both (with willingness to work in both)

• Experience with SDN or cloud networking systems (VPCs, virtual routers, SDN controllers, overlay networks)

• Strong understanding of IPv6, routing (BGP/ECMP), and common overlay technologies (e.g., VXLAN/Geneve)

• Hands-on experience with DPDK, VPP, eBPF, or XDP

• Strong debugging skills across distributed systems and multi-layer networking

Preferred

• Experience building high-performance proxies, L4/L7 load balancers, or edge networking components

• Experience building or operating multi-tenant cloud networking systems

• Familiarity with distributed systems fundamentals (consensus, sharding, leader election, etc.)

• Experience with traffic engineering or global traffic routing

• Security background in network isolation, segmentation, or zero-trust architectures

Bonus

Experience with any of the following is a plus:

• Kernel bypass / SmartNIC / SR-IOV / RDMA

• Custom BPF programs for observability or dataplane filtering

• Large-scale IPv6-only environments

• Advanced traffic simulation or network modeling

• Ultra-low-latency performance tuning

• Cloud provider networking (AWS VPC, GCP VPC, Azure VNet internals)

About Blaxel

Blaxel is AWS for AI agents. We’re building a new kind of cloud optimized for agentic AI—powered by a purpose-built 25ms cold-start serverless compute engine.

We process billions of AI agent requests and power the production infrastructure for top AI startups. Traditional clouds weren’t built for agentic workloads—we’re solving the performance, reliability, and stateful execution challenges they can’t.

We recently raised a $7.3M seed round led by First Round Capital.