OVERVIEWJob Type
Full-time
Location
Springfield, VA 22151 US (Primary)
Job Description
ITPIE ships as a hardened appliance: a GPU server installed at the customer's site that runs ITPIE's applications without an internet connection. You are ITPIE's engineer at that site. You install the appliance, keep it running, and respond first when something breaks.
You fix what you can on site. When you can't, you collect the right evidence and send it to the engineering team. On a disconnected network, that evidence is often all they get, so its quality decides how fast the customer gets a fix. This is not a scripted help-desk role: you must understand the appliance well enough to reason about failures no procedure covers.
VAE operates across three lines of business: professional engineering services, value-added resale of IT products, and development of the ITPIE portfolio of software applications. The ITPIE portfolio is sold primarily, but not exclusively, to U.S. government customers. VAE has invested significant internal resources in the development of ITPIE for nearly 15 years.
WHAT YOU WILL WORK WITH
- Hardware. Enterprise rack-mount GPU server with BMC, hardware RAID, and GPUs passed through to VMs.
- Platform. Rocky Linux hardened to the DISA STIG (SELinux enforcing, FIPS), HashiCorp Nomad, Podman containers and QEMU/KVM virtual machines, delivered as signed offline bundles with A/B updates.
- Services. Traefik, Keycloak (password and CAC sign-in), PostgreSQL, Redis, Grafana, and ITPIE's own applications.
- Management. A web interface and a console application for most tasks, and a root shell for the procedures in the Quick Reference Handbook.
- Network. Air-gapped or restricted, with no host firewall: the site's network controls are part of the security design.
CORE RESPONSIBILITIES
- Build and rehearse. Build each customer's appliance before it ships, install its workload bundle, and rehearse the deployment at Springfield with the platform team.
- Install and accept. Rack, power and network the appliance with site staff; bring it to its STIG-assessed configuration; set up BMC, time, DNS, TLS certificates and the customer's PKI.
- Operate. Deploy updates and roll back when needed; watch disks, volumes, workloads and log forwarding; manage accounts and sign-in; take off-box backups when on-box rollback won't fit.
- Protect and assess. Keep custody of the disk-encryption recovery passphrase, restore unattended boot after a TPM or motherboard change, and give the customer's assessors evidence for the controls the site owns.
- Diagnose. Find the cause across the full stack: an SELinux denial, noexec mount, systemd sandbox setting or UID mismatch behind a "permission denied"; a Nomad job that won't place, won't become healthy, or keeps restarting; a server that won't boot, loops, or falls back to the old slot after an update; a VM that can't use its GPU or network; a CAC sign-in that fails on the certificate chain or CRL.
- Escalate. Collect and redact the evidence (journal, audit and AVC records, Nomad, Podman and QEMU state, boot loader state, BMC logs and hardware status), send it through the approved channel, and write a clear report of what you saw, what you did, what changed, and what you think the cause is.
- Improve the product. Tell engineering when procedures or docs are wrong. Install and verify fixes at the site.
Qualifications
WHAT YOU BRING
- 3+ years of production Linux administration, preferably RHEL or Rocky.
- Strong systemd, journald, shell, LVM and NetworkManager skills.
- Hardened-system experience: SELinux enforcing, DISA STIG, FIPS mode. You can read an AVC denial and explain it.
- Containers (Podman or Docker) and KVM virtual machines.
- Enterprise server hardware: BMC/Redfish, hardware RAID, disk replacement, UEFI.
- Evidence-driven troubleshooting, and clear writing for engineers who aren't on site.
- Active Secret clearance required; Top Secret preferred. U.S. citizenship required.
- Willingness to work on site, sometimes in a SCIF or on a closed network.
Nice to have
- HashiCorp Nomad or Kubernetes.
- Secure Boot, TPM 2.0, LUKS2, GRUB; OpenSCAP, auditd, RMF.
- PKI, CAC, Keycloak, Traefik, PostgreSQL.
- Data-center GPUs and PCI passthrough.
- Reading Go: the appliance's own tools are written in it.
- NIPRNet/SIPRNet experience; a DoD 8140 certification such as Security+ or RHCSA.
HOW YOU WORK WITH THE TEAM: You are part of the engineering team, not a separate support group. You have the design docs and the source, and you join the engineering discussions about the failures you find.
THE DETAILS
- Location: customer site [TBD: city], with training and rehearsal time in Springfield, VA.
- Travel: about 30%, including international (CONUS and OCONUS).
- Clearance: Active Secret clearance required; Top Secret preferred. U.S. citizenship required.
- Compensation: [TBD], commensurate with experience.
- Employment type: full-time.
Clearance Level
Secret
Certifications
8570 compliant
Salary Range
$115,000 - $145,000