← All roles

Staff Engineer, Software Distribution onPrem

Qevlar AI· Qevlar AI
Apply on company site
Hybrid (Paris)ContractPosted 5d ago

About the role

Staff Engineer responsible for delivering on-premise production deployments for Qevlar AI's software, including architecture scoping, packaging, self-hosted infrastructure, and end-to-end delivery in air-gapped environments.

Responsibilities

  • Decide what we do not port. Define on-prem reference architecture and dependency list.
  • Ship the bundle: Helm chart and air-gapped tarball installable on vanilla Kubernetes and OpenShift without internet access.
  • Replace the cloud with self-hosted equivalents (Postgres, S3-like storage, secret management, identity federation).
  • Package and size the inference layer: manage serving stack, VRAM/concurrency sizing, and distribution of model weights in air gaps.
  • Build the test harness to make releases boring: ephemeral environments, install/upgrade smoke tests across Kubernetes distributions, identity providers, network modes.
  • Ship clean: signed images, SBOMs, minimal base images, no critical/high CVEs at release.
  • Impose upgrade contracts across the organization: backward-compatible migrations, feature flags, health checks, and support bundles.
  • Keep the product simple and veto problematic dependencies.

Qualifications

  • Shipped software that is installed by others; experience operating your own cluster is essential.
  • Deep Helm expertise (subcharts, conditional templating, values design), OCI registries, and broad Kubernetes experience (OpenShift, vanilla Kubernetes, RKE2, k0s).
  • Experience in air-gapped/heavily restricted environments (image mirroring, offline bundles, egress proxies, custom CA bundles).
  • Replaced managed cloud services with self-hosted equivalents (Postgres operators, MinIO/S3-compatible storage, Vault/External Secrets).
  • Software supply chain security: image signing, SBOM generation, distroless/minimal base images, CVE triage.
  • Strong IaC and CI skills (Terraform, Ansible, pipelines).
  • Hands-on experience serving models on GPU (vLLM, NVIDIA NIM, TensorRT-LLM) including VRAM/concurrency sizing and NVIDIA GPU Operator/Container Toolkit.
  • Strong Go and/or Python coding ability.
  • Ability to exercise architectural authority and say no when needed.
  • Fluent in French and English.

Skills

Required:Shipped software that others installHelmKubernetes (OpenShift, vanilla, RKE2, k0s)Air-gapped / restricted environmentsSelf-hosted equivalents (Postgres, MinIO, Vault/External Secrets)Software supply chain securityTerraform / Ansible / CIGPU model serving (vLLM, NVIDIA NIM, TensorRT-LLM)Go / PythonFrench and English fluency

Benefits

  • Equity and early ownership
  • Autonomy and creativity
  • Meaningful, real-world impact in cybersecurity
  • Flexible location: remote or Paris office
  • International, fast-moving team environment
Last updated Oct 21, 2026. Interested? Apply on the company site.

About Qevlar AI

Let your SOC team focus on real threats. Not alert firefighting.

Qevlar AI | Autonomous SOC Platform

AIAutonomous SOCDetection EngineeringIncident ResponseIncident responseSecurity OperationsSecurity Operations CenterThreat Hunting
Company profileWebsite

More open roles at Qevlar AI