Site Reliability Engineer, Strategic Defense

Title: Site Reliability Engineer, Strategic Defense Location: Washington, District of Columbia, United States

ABOUT THE TEAM

The Strategic Defense team at Anduril protects the United States and its allies from ballistic, hypersonic, and air-breathing threats. Those threats now arrive in seconds, maneuver mid-flight, and come in volleys designed to overwhelm the legacy systems meant to stop them — so we're building a software-defined command and control layer that can see, decide, and act faster than any human operator could alone. Our team spans autonomy, tracking, simulation and wargaming, hardware, systems engineering, and test and evaluation to build an accredited fire control system fielded rapidly and at scale. We are at the forefront of incubating and maturing AI-native offerings for the integrated air and missile defense domain.

ABOUT THE JOB

As an SRE in Strategic Defense You’ll rack and provision hardware, build the automation that turns a blank server into a hardened node in a running cluster, design how software reaches systems that are only sometimes connected, and debug failures from the Linux kernel up through the network to the pod. You’ll work closely with software, data and operations engineers to get systems into the hands of customers, deployed and supported on real infrastructure with real users.

Above all, you bring a “Whatever It Takes” mindset: pragmatic, fast, and focused on the mission, with sound engineering judgment about what to build now and what to build next.

WHAT YOU’LL DO

• Provision bare metal at scale. Run the pipeline that takes servers from rack to production: PXE boot, kickstart, BMC/iDRAC, firmware, STIG hardening, and handoff into Kubernetes.
• Run Kubernetes without a managed control plane. Operate self-managed clusters (RKE2, k3s, Talos) on-prem and at the edge: networking, storage, upgrades and capacity.
• Design for disconnected environments. Decide how software and data reach hundreds of nodes that may be offline, with limited bandwidth and short update windows.
• Move software across network boundaries. Ship and update systems on air-gapped and classified networks, including through cross-domain solutions and data diodes.
• Debug at every layer. Trace failures across Linux, the network and Kubernetes, then lead the root cause analysis and postmortem.
• Automate the toil. Write the Python and Bash tooling the fleet runs on, and manage infrastructure as code.
• Deploy with customers. Travel to sites to install and support systems, often as the only engineer there.

REQUIRED QUALIFICATIONS

• Currently possesses and is able to maintain an active U.S. Top Secret security clearance.
• 3-8+ years of hands-on infrastructure, systems or site reliability engineering.
• You can explain Linux process states, system calls, memory, file systems and inodes, systemd, and permissions/SELinux, and you can find the process behind a symptom from the shell without a cloud console.
• You network using first principles: TCP/UDP, DHCP, DNS, routing, VLANs and firewalls. You can trace a request hop by hop from client to pod and name the tool you’d use at each step.
• You know Kubernetes below the managed layer: how Services, kube-proxy, CNI, DNS, scheduling, requests/limits and storage actually work, and how to debug them with kubectl, not a UI.
• You write working Python and Bash from scratch, without looking things up: parsing logs and text, handling files and processes, adding retries and error handling, and reading and fixing someone else’s script.

PREFERRED QUALIFICATIONS

• Experience building, deploying or operating systems on air-gapped, classified or security-hardened networks.
• Experience designing distributed systems for the edge or for disconnected, intermittent, and limited (DIL) conditions: mesh networks, store-and-forward, peer distribution, QoS.
• Infrastructure as code and GitOps (Terraform, Ansible, ArgoCD/Flux); experience with Nix is a plus.
• Observability stacks (Prometheus, Grafana, Loki, OpenTelemetry) and turning their signals into decisions about alerts and capacity.
• Platform and supply-chain security: image signing, PKI and certificate authorities, disk encryption (LUKS), FIPS.
• Ability to read, debug and modify software written in Go, Python, Rust or C++.