Services / DevOps and platform engineering
DevOpsWe design and operate the pipeline, GitOps, and cluster for your application
We design, deploy, and operate a software delivery platform. We start from the application: repositories, tests, secrets, environments, and who deploys. Then we stand up a pipeline in GitLab or GitHub Actions, Kubernetes, and Argo CD. We build the application in CI and deploy it from Git, staging first, then production, with a dashboard and on-call.
How it fits together
From the application to the pipeline, the cluster, and on-call
First we understand the application and how the team deploys. On that basis we put together CI, Kubernetes, Argo CD, and monitoring. GitLab, GitHub Actions, and Grafana are what we use for that work every day.
- Preview and production use the same manifests, with different configuration.
- We record a cluster change in Git. A manual fix in production goes back into the repository.
We build a pipeline that tests and ships the application
In GitLab or GitHub Actions we put together the test, scan, image, and a gate to the environment. A merge goes through the same checks. Preview comes up on the merge request and disappears after close.
Gates
Tests, lint, SAST, and image policy. Only what can be deployed lands on main.
Environments
Dev, staging, and production on the same pipeline. The difference is approval and secrets.
Identity
Runner, registry, and cloud on service accounts, with a short-lived token.
Audit trail
Who ran the job, which commit, which artifact went further. That is material for a change audit.
We deploy the application from Git, staging first
The pipeline builds the image. Argo CD compares the cluster with the repository and deploys what passed review. Staging first, then production. Rollback is a return to a known commit.
Source
Manifests, Helm, or Kustomize in the repository. Production is reached by merge.
Visibility
Health, sync, and diff on the application. OutOfSync has an owner.
Boundaries
A project per team or per client. Production has narrower access.
Failure
When sync fails, you see the resource and the event. A playbook for rollback and escalation.
We stand up a cluster the team can deploy to
Kubernetes in the cloud or in the server room, with the same network, limit, and copy policy. Node, storage, and ingress are written in code. A production change goes through Git. The dashboard is for diagnosis.
Repeatability
Terraform for the cluster, add-ons, and network. A second environment comes up from the same description.
Limits
Request, limit, quota, and NetworkPolicy. One workload does not take CPU from the others.
Operations
Upgrade, certificates, storage class, and PDB. The change window and rollback plan are in the scope.
Access
kubectl through corporate identity. Every API request has a trail.
On-call sees the application after every deploy
Metrics, logs, and traces go to Grafana. A dashboard per service. An alert has a runbook. After a deploy you see whether p95 and errors stayed in the SLO budget.
Signal
RED or USE per service. Logs with pod and commit context. Tracing where the path is distributed.
Alert
Priority, silence during a maintenance window, escalation. We cut noise in the first month.
Cost
We set retention and sampling on purpose, so logs do not blow the bill.
People
The team and on-call see the same dashboard.
How we work
From the client application to production
We start from the application: repositories, tests, secrets, environments, and who deploys. Then we stand up the pipeline, the cluster, and Argo CD. We build the application in CI and deploy it from Git, staging first, then production. We stay for dashboards, SLO, and on-call.
- Application We check repositories, the stack, secrets, and who deploys today. We agree environments. That gives the scope: pipeline, cluster, GitOps.
- Pipeline In GitLab or GitHub Actions we put together the test, scan, image, and gates. Preview comes up on the merge request.
- Cluster and staging We stand up Kubernetes, limits, and Argo CD. Staging has the same shape as production, only smaller.
- Production and on-call Argo CD deploys what passed review. After go-live: dashboard, SLO, alert, and a drift review.
Talk
We will discuss the application, the pipeline, and deploys
On that basis we will prepare a scope: CI, cluster, GitOps, and on-call.
Contact us