All services

Cloud & Platform Engineering

Releases you can trust, environments that match and rollbacks you have already rehearsed.

A release pipeline. A reviewed commit is built and tested in CI, packaged once as a versioned container image, and deployed unchanged to staging and then production, both defined as infrastructure as code. Production can roll back to the last release. Monitoring, alerts and runbooks cover every environment.Rollbackv2.4.0CommitReviewed changeBuild and testEvery commit, in CIContainer imageBuilt once, versionedStagingProductionSame image everywhereInfra as codeLast releaseMonitoring and alertsRunbooks for on-callA release pipeline. A reviewed commit is built and tested in CI, packaged once as a versioned container image, and deployed unchanged to staging and then production, both defined as infrastructure as code. Production can roll back to the last release. Monitoring, alerts and runbooks cover every environment.Rollbackv2.4.0CommitReviewed changeBuild and testEvery commitContainer imageBuilt onceStagingProductionSame imageeverywhereLast releaseInfra as codeMonitoring and alertsRunbooks for on-call
A release pipeline we’d set up. The highlighted line is one release, from a reviewed commit to production.

When teams bring us in

  • Releases are manual or risky and need heroics every time.
  • Environments have drifted apart, so what works in staging fails in production.
  • Outages take too long to diagnose.
  • Workloads are moving to containers and Kubernetes.

What the work covers

CI/CD pipelines

Consistent build, test and validation stages for every change, so a release is the same process every time.

Release versioning and rollback

Clear versioning, promotion rules between environments and rollback paths that are defined before anyone needs them.

Environment parity

Development, staging and production built the same way, so surprises don’t wait until release day.

Containers and orchestration

Docker, Kubernetes and OpenShift, set up so your own team can operate them.

Cloud and infrastructure as code

AWS and Azure infrastructure described in code, reviewed like any other change and repeatable.

Monitoring, alerting and runbooks

Dashboards and alerts that point to the cause quickly when something breaks, and runbooks that say what to do next.

What you get at the end

  • Pipelines you can trust
  • Documented infrastructure you can rebuild the same way every time
  • Rollback procedures that have been rehearsed
  • Dashboards and alerts
  • Runbooks for whoever is on call

How an engagement runs

  1. Discovery

    We look at how releases happen today, where they fail and what slows them down, then agree on the order of the improvements.

  2. Harden in small steps

    We automate the recurring steps, add guardrails and verification checks, and define rollback paths, without stopping your team from shipping.

  3. Rehearse and hand over

    We rehearse the rollbacks, write the runbooks and walk your team through them, so the next incident isn’t the first time anyone tries.

Thiago Benkendorf, the senior software architect and engineer behind Nithron, does every engagement hands-on. On bigger projects he can coordinate other engineers, as he did with a six-person contractor team on a manufacturing platform rebuild, and he stays your single point of contact.

Technologies

  • Docker, Kubernetes and OpenShift
  • AWS and Azure
  • CI/CD pipelines and infrastructure as code
  • Monitoring and alerting

Questions

Do we have to move to Kubernetes?

No. The operational side can start simple with container-based deployments and grow from there. We recommend Kubernetes when it solves a problem you actually have.

Which clouds do you work with?

AWS and Azure, as well as OpenShift.

Can you improve our releases without a release freeze?

Yes. We change the pipeline in small, safe steps, so your team keeps shipping while releases get safer.

Can you be on call for us?

We can, as part of ongoing support. Either way, you get runbooks so your own team can handle the common incidents.