Network Engineer, Autonomous Factory

Foundry Robotics•San Francisco, CA

3d•$150,000 - $250,000•Onsite

About The Position

Foundry Robotics is building an AI-native robotics manufacturing company focused on deploying advanced assembly and production capability for leading robotics companies and national-security-critical hardware. Factory OS is building software that runs factories. Our platform spans cloud backends, on-prem services, and edge devices deployed on the factory floor. The operators, supervisors, and managers who use our tools need interfaces that are fast, reliable, and intuitive—even when network conditions aren’t perfect and the environment is demanding. As a Network Engineer, Autonomous Factory, you’ll design and build that network from the ground up for a 130,000 sq. ft. autonomous manufacturing facility, and then own it as we scale. You’ll work across the full network stack: physical infrastructure, switching and routing, server and GPU rack fabric, segmentation between industrial and IT systems, wireless, and core services including DHCP, DNS, and time synchronization. You’ll also own the WAN and VPN edge, the overlay network engineers use to securely reach the factory floor, and the security boundaries required for ITAR and CMMC 2.0 Level 2 compliance. This is a deeply hands-on role. You’ll be in the ceiling, in the racks, and in the packet captures, while also working with engineering and manufacturing teams to design how new robotic work cells should be connected before the first one is ever built. If you want to build critical infrastructure from the ground up and have a direct role in enabling a truly autonomous factory, this is an opportunity to make a meaningful impact from day one.

Requirements

7+ years in network engineering, including at least one greenfield or major build-out of a campus, data center, or industrial facility.
Deep routing and switching: OSPF or BGP, VLANs and VXLAN, spanning tree, LACP, QoS, multicast, and the judgment to know when not to use them.
Physical layer competence: fiber and copper plant design, cabling standards, rack and pathway planning, working with electrical and facilities trades.
Firewalls, VPN (IPsec, WireGuard), and zone-based segmentation. You have written and defended a security boundary.
Data center networking for compute racks: 25/100G top-of-rack, storage and GPU traffic patterns, out-of-band management.
Comfortable on Linux with tcpdump, iproute2, and nftables. You can debug a problem that lives between a host's routing table and a switch.
Able to document and automate: network as code, version-controlled configs, and monitoring you built yourself.

Nice To Haves

Industrial or OT networking: PLCs, EtherCAT or PROFINET, robot controllers, safety systems, IEC 62443 concepts.
ROS 2 / DDS on real networks, including multicast discovery.
WireGuard-based overlays at scale and the underlay/overlay interaction problems that come with them.
AWS networking: VPC, Transit Gateway, Site-to-Site VPN or Direct Connect, private DNS.
Video and camera networks with high camera counts and bandwidth planning.
PTP or other precision time sync for robotics or measurement.
Prior work under ITAR, CMMC, or NIST SP 800-171, ideally producing assessment evidence.

Responsibilities

Produce the network design for the factory: physical plant, logical topology, IP plan, segmentation model, redundancy, and capacity. Get it built with contractors and facilities.
Design and stand up the server and GPU rack fabric: high-speed top-of-rack, out-of-band management, storage traffic, and coordination with power and cooling.
Wire and segment robotic work cells: controllers, safety systems, fieldbus segments, edge compute, cameras, and operator stations. Make robotics middleware discovery and multicast behave.
Own the WAN and VPN edge: redundant private connectivity to the cloud that actually fails over, with monitoring that catches lost redundancy before it becomes an outage.
Run the overlay network and its interaction with the underlay. Design so overlay routing can never hijack local traffic.
Operate the core services the floor depends on: DHCP, DNS including private zones into the cloud, NTP or PTP, network access control where it fits, and firewall policy between zones.
Instrument everything. Switch, firewall, and flow telemetry land in the shared monitoring stack, and you write the runbooks.
Carry the network portion of ITAR and CMMC 2.0 Level 2: boundary protection, access control, logging, configuration management, and evidence for assessors.
Enable remote fleet operations: re-imaging, health checks, and access to cell compute without walking to the cell.