Senior Release Engineer

Company Overview

Open Innovation AI is a global technology company that specializes in developing advanced solutions for managing AI workloads. Its flagship product, the Open Innovation Cluster Manager (OICM), orchestrates complex AI tasks efficiently across diverse infrastructures. The platform is hardware-agnostic, optimized for various GPUs and accelerators hardware, and facilitates seamless integration and scalability for enterprise AI applications. Open Innovation AI focuses on optimizing and simplifying AI workload management and making AI technologies accessible to organizations of all sizes. With its innovative solutions, companies can reduce operational costs, accelerate time to value, and maximize their return on investment, ensuring that their AI strategies contribute directly to enhanced business outcomes.

Role Overview:

As Senior Release Engineer, you own how the platform goes from source code to a validated, shippable product. That means the full air-gapped build pipeline: golden operating-system images built with Packer, the installer ISO, the container image and Helm chart bundles, the offline package and registry mirrors, and the Ansible execution environment that ties them together. When a release is cut, it comes out of your pipeline.

You also own the automated release-validation gate: an end-to-end test harness that stands up a full cluster from bare metal and exercises an in-place upgrade before any build reaches a customer. Building this gate is the first and most valuable thing this role delivers, and it directly lowers the cost and risk of every deployment.

This is a senior individual-contributor role. You design the build and test systems as maintainable software, not throwaway scripts, and you lead through the pipelines you build, the quality gates you set, and your judgment about what is safe to ship. You work directly with the Head of Infrastructure, you codify release-cutting into a repeatable, documented process, and you grow the release-engineering practice as the platform and the team expand.

This is an internal engineering role. You build the product; there is no customer travel or on-site work.

Role Responsibilities:

• Own the air-gapped build pipeline end to end: Packer-built, CIS-hardened golden OS images, the installer ISO, container image and Helm chart bundles, offline package (apt/dpkg) and registry mirrors, and the Ansible execution environment.
• Build and own the automated release-validation gate: an end-to-end harness that provisions a full cluster from bare metal and exercises an in-place upgrade on every release candidate.
• Cut and package releases: produce the versioned, self-contained, fully offline artifact that ships to customers.
• Own the CI that builds, tests, and validates the platform on every change.
• Own the security-patch and CVE-response pipeline: rebuild, revalidate, and republish affected artifacts quickly and safely.
• Maintain the offline registry cache and artifact server that feed both the build and the deployed clusters.
• Keep the build reproducible and artifacts self-contained: every image, chart, and package consumed at deploy time comes from the bundle, never the internet.
• Document the pipeline and its operations: write the ADRs, build runbooks, and release procedures that let another engineer cut, validate, and troubleshoot a release without you.
• Work with the Head of Infrastructure to harden the release process against regressions and unsafe upgrades.

Required experience & Qualification

• Strong scripting and automation: fluent in Bash and Python, building reusable, maintainable tooling rather than throwaway scripts.
• Strong software-architecture judgment: you design pipelines, tooling, and test harnesses as coherent, maintainable systems.
• Deep Linux systems expertise: packaging (apt/dpkg), OS image building (Packer or equivalent), system services, and troubleshooting.
• Strong with containers: building, inspecting, and managing Docker/OCI images and containers.
• Strong with Helm: chart packaging, templating, and reconciliation.
• Deep ownership of the release process: versioning, packaging, and CI/CD, cutting reproducible releases that ship as deployable artifacts rather than hosted services.
• Strong end-to-end and integration testing: building automated harnesses that provision and validate real infrastructure.
• Strong infrastructure-as-code discipline, with hands-on Ansible (this platform is Ansible-driven).
• Working knowledge of Kubernetes and the mechanics of installing and upgrading it, since you will validate cluster installs and upgrades.
• Experience with air-gapped, offline, or otherwise constrained delivery (registry mirrors, package mirrors, self-contained bundles), or a clear track record that shows you can pick it up.
• A bias for reproducibility, idempotency, and fail-loud validation.
• Fluent written and spoken English.
• Deep storage, networking, or GPU expertise is not required. You will validate that those subsystems come up correctly, not architect them

Preferred Qualifications

• Packaging or validating GPU or RDMA-enabled Kubernetes stacks.
• Software supply-chain security: artifact signing, provenance, and SBOMs.
• Prior work on a self-hosted or on-premises product distribution.

What Success Looks Like in the First 90 Days

• The release-validation gate runs automatically on every release candidate: a full green-field install and an in-place upgrade, so a release ships straight from the automated gate.
• Releases are cut from your pipeline as a repeatable process you fully own.
• Deployment-blocking issues are caught by the gate before a build reaches a customer, not in the field.
• The build is reproducible: the same source produces the same self-contained, offline artifact every time.
• The pipeline and its runbooks are documented well enough that another engineer can cut and validate a release from them, so shipping is not tied to any single person.