Confirmed on the employer's own hiring board on Nov 20, 2026. First seen by Alion on Nov 19, 2026.
About the role:
As a Cloud DevOps Engineer at Lumafield, you will own the cloud platform that runs Voyager, our CT data platform. A scan is only useful once someone can open it, measure it, and make a decision, and that happens in Voyager, in a web browser. Everything underneath is yours: AWS, the production Kubernetes clusters, Terraform, the pipeline every engineer ships through, observability, and the cloud side of our security and compliance programs.
This is an ownership role, not a ticket queue. Nobody will hand you a backlog; you will be the person who knows what the platform needs, what it costs, and what to fix first. We expect you to grow into a senior engineer here.
You will work in Platform Engineering alongside the backend lead, the software engineers who ship through your pipeline every day, and the team building the on-prem version of Voyager for customers who cannot send scan data to the cloud. You will report to engineering leadership; Platform Engineering is part of the organization our new Head of Software will lead.
This is a hands-on, in-person role based in our San Francisco, CA office.
What you'll do:
- Own our AWS accounts and regions: architecture, access, onboarding and offboarding, and a bill you can explain by service and environment, then bring down
- Run the production EKS clusters, and carry Kubernetes version upgrades from plan and rehearsal to completion, with a rollback ready
- Keep infrastructure and Kubernetes configuration in Terraform, so changes go through code review instead of the console
- Own CI/CD - GitHub Actions, CI runners, artifact and image storage, and GitOps deploys with ArgoCD - and make it faster and more reliable in ways engineers notice
- Own observability (Sentry, Logz.io and Postman today) and decide what the platform needs next; take on-call, and close every incident with a written postmortem and a fix
- Own the cloud side of our SOC 2 Type II and ISO/IEC 27001:2022 programs: least-privilege access, logging, encryption, cluster hardening, and audit evidence that comes out of the system instead of being assembled by hand
- Partner with our IT team on shared controls, and raise the platform's security bar as we maintain CMMC Level 1 and ITAR compliance and work toward FedRAMP 20x
- Document the platform - runbooks, postmortems, and design notes - so it does not live in one person's head
About you:
- You have been on the hook for a production platform: releases, incidents, and on-call
- You have upgraded a production Kubernetes cluster and can tell us what went wrong and what you changed afterwards
- You write Terraform that other people depend on, and you are comfortable with modules, state, and reading a plan before you apply it
- You have built or rebuilt a CI/CD pipeline that engineers used daily
- You have opened a cloud bill, found what was driving it, and changed it
- You have implemented a security control or produced audit evidence for a program such as SOC 2 or ISO 27001, or done the equivalent access and hardening work without the audit
- You write clearly: runbooks, postmortems, and design notes someone else can act on
- Production experience on ECS rather than Kubernetes, or on GCP/GKE rather than AWS, still counts when your infrastructure-as-code work is strong; tell us how you would close the gap
Bonus:
- Large-file or 3D data pipelines, or GPU workloads
- Software that talks to hardware in the field
- Regulated environments such as FedRAMP, ITAR, or CMMC

