How the platform works.
OpsWise is your entire DevOps team, plus the infrastructure it runs on. Instead of hiring engineers and piecing together a dozen tools, you get a complete, preconfigured platform — hosted on your own off-cloud servers and operated by a team with a NOC behind it. You build the product; we run the platform.
The core technology.
At the core sits Incus — the open-source hypervisor and container manager — combined with Linstor, a software-defined block storage layer that replicates your data across nodes.
Incus runs the virtual machines and containers; Linstor keeps the storage safe and synchronized. Together they give you a private cloud that behaves like a public one — without the lock-in or the bill.
Why this stack.
The stack is the result of years of hard-won experience. It was chosen by DevOps engineers who have run production systems for years, and every tool earned its place, not because it’s trendy, but because it’s the optimum, proven combination a tech company needs to build software or a digital product. No experiments, no dead ends.
Architecture & reliability.
Single sign-on is included by default — one login for Git, CI, Kubernetes, docs, and chat, via Keycloak.
The hardware architecture is built to survive failure. It’s designed so the cluster always withstands losing one third of its nodes, and data is replicated via Linstor — so when a server dies, nothing is lost and the platform keeps running.
No vendor lock-in.
You’re never locked in. Everything is open source and runs on servers you rent or own. Beyond the stack you choose, you can spin up your own VMs for your own Kubernetes clusters, or any workload you need — staging, GPU boxes, a client’s project. It’s your private cloud, no strings attached.
Migration, made easy.
The chosen technologies come preconfigured and wired together, so moving from your current setup is measured in days, not months — and we do the heavy lifting.
The team behind it.
We’re a DevOps team with a NOC. We replace classic DevOps positions: monitoring, maintenance, upgrades, and SLA-backed support are handled by us, around the clock. You get an experienced team instead of hiring one — or several — engineers.
The tools, in detail.
CI/CD & development
From the first commit to a live deployment, the whole pipeline runs on your own servers.
Gitea or GitLab
Your Git host. Gitea is lightweight and fast; GitLab is the full-featured option when you need more. Your repositories, issues, and merge requests stay on hardware you control.
Jenkins + runners
Automation for builds, tests, and deployments. Runners execute your jobs in isolated environments, and scale with the load you actually have.
ArgoCD
GitOps deployments. Your cluster continuously matches what’s defined in Git, so every change is versioned, reviewable, and reversible.
Harbor
A private container registry with vulnerability scanning and fine-grained access control. Your images stay private, on your infrastructure.
k3s
Lightweight, production-ready Kubernetes. One click and you have a full cluster — without the operational overhead of managing it yourself.
MinIO
S3-compatible object storage. Drop-in replacement for cloud storage buckets — used for artifacts, backups, and anything else that needs an object store.
Quality
SonarQube
Continuous code analysis and security scanning. Bugs, code smells, and vulnerabilities are flagged on every build, before they reach production.
Docs & chat
Knowledge and communication, self-hosted and private.
Docmost
Team documentation and wikis — a Notion alternative. Everything your team writes lives on your own servers.
OnlyOffice
A full office suite (documents, spreadsheets, presentations) — a Google Docs alternative. Collaborative editing without handing your files to a third party.
Mattermost
Team chat — a Slack alternative. Channels, threads, and integrations, self-hosted and under your control.
Observability
Metrics, logs, and traces — everything you need to see what’s happening, in one place.
Grafana
Dashboards and visualizations over all your data sources. One place to watch the health of your platform.
Prometheus
Time-series metrics collection with alerting. Know the moment something drifts from normal.
Tempo
Distributed tracing. Follow a request through every service to find exactly where the latency is.
Loki
Log aggregation, tightly integrated with Grafana. Search and correlate logs alongside your metrics.
Graylog
Centralized log management with powerful search and alerting, for when you need deep log analysis.
Uptime Kuma
A status page and uptime monitor. Get pinged the moment a service goes down — and share a public status page with your users.
Security & access
Keycloak
Single sign-on for the entire platform. One login, one identity, one place to manage who can access what.
WireGuard
A fast, modern VPN. Access your private infrastructure securely from anywhere, without exposing it to the public internet.
AI & automation
Ollama
Self-hosted large language models and AI. Run assistants and tools on your own hardware, keeping your data private.
MCP server
Your platform is exposed to AI agents through the Model Context Protocol, so your agents can handle the basic, repetitive tasks for you — checking CI status, creating issues, triggering deployments.