About Orasio
Orasio builds cutting-edge solutions that automatically detect situations of interest in video streams. Powered by advanced artificial intelligence models trained to extract actionable insights from images, our technology can be deployed across existing IT systems or new infrastructures—for private companies or public institutions, civilian or military—while strictly complying with applicable legal frameworks.
As AI becomes a key pillar of technological sovereignty, Orasio aims to be Europe’s leading player in video intelligence. Our mission is to develop world-class AI solutions tailored for defense and public safety sectors, combining technical excellence with unwavering respect for personal data protection laws across jurisdictions.
Your Mission
We're looking for a Senior DevOps / SRE Engineer to make our video intelligence platform observable and reliable wherever it runs, with a clear focus: enhance public safety through AI.
Our platform doesn't only live in the cloud. It runs on Kubernetes clusters in physical datacenters, on customer premises, on GPU servers and edge devices, and in fully air-gapped sites with no internet access at all. You will own how we monitor and alert on all of it, so that every cluster, node, GPU and video pipeline is watched and every real problem reaches the right person, whether the site is connected or not.
Your Responsibilities
Own our observability stack (metrics, logs, dashboards, alerting) across every Kubernetes cluster, from cloud to on-prem to air-gapped
Design meaningful alerts and routing (SLOs, severities, on-call), so that alerts lead to action and not to noise
Monitor the full stack: Kubernetes, physical hosts, NVIDIA GPUs, storage, network and our video and AI workloads
Build monitoring that works with no internet access: offline install kits, local alerting, and a way to bring signals back from disconnected sites
Deploy and operate Kubernetes (k3s) clusters on bare metal and in physical datacenters, alongside managed clusters in the cloud
Design and troubleshoot on-prem networks: VLANs, routing, firewalls, DNS, load balancing and high-throughput links for video
Run our GitOps delivery with Argo CD across a hub-and-spoke fleet of clusters
Automate provisioning and configuration with Ansible, OpenTofu and Helm, and keep everything reproducible
Work with Software, AI and Deployment engineers to make their services observable from day one
Lead incident response and post-mortems, and write the runbooks that let anyone on the team act on an alert
What We’re Looking For
5+ years of experience in DevOps, SRE or infrastructure engineering, with at least 3+ years running Kubernetes in production
Deep hands-on experience with a Prometheus-compatible monitoring stack (Prometheus or VictoriaMetrics, Alertmanager, Grafana) and log aggregation (Loki, VictoriaLogs or similar)
A track record of designing alerting that people trust: clear thresholds, SLOs, routing and on-call practices
Real experience with on-prem infrastructure: bare-metal servers, physical data centers, hardware failures and the constraints of sites you can't reach in a few clicks
Strong networking fundamentals: TCP/IP, VLANs, routing, firewalls, DNS, VPNs and troubleshooting across network boundaries
Experience with air-gapped or restricted environments: private registries, mirrored artifacts, offline installs and upgrades
Infrastructure as code and GitOps: Helm, Ansible, Terraform/OpenTofu and Argo CD or Flux
Strong Linux skills, with an eye for performance, reliability and security hardening
Ability to work independently on complex, multi-site issues and to explain them clearly to the team
Nice to have
Experience with the VictoriaMetrics ecosystem (vmsingle, vmalert, VictoriaLogs) and Grafana Alloy
GPU monitoring and operations: NVIDIA drivers, GPU Operator, DCGM exporter
ML lifecycle and model delivery: versioning, packaging and releasing AI models alongside the services that use them, including to offline…