Cloud infrastructure & DevOps
Codified, deployable, observable, and costing what it should — the platform under your product, from senior engineers who have operated production at scale.
Last updated:Cloud infrastructure and DevOps consultancy covers the platform an application runs on: infrastructure as code, deployment pipelines, observability, and cloud spend. Revenant Systems builds and reviews cloud infrastructure with Terraform and Terragrunt on AWS, Azure, and Google Cloud, and delivers to Kubernetes through ArgoCD where that suits the team and the product.
Who is this cloud service for?
Teams running production on AWS, Azure, or Google Cloud whose platform has become the constraint — deployments feared, spend unexplained, or an estate one hand-built console away from unrecoverable. It suits companies with engineers who own the product but nobody who owns the platform. A team choosing its first hosting plan does not need this yet.
Where does cloud work start?
Cost problems start with the fixed-price cloud cost audit, from £4,950 — spend is where evidence is cheapest to gather, and the findings often fund the rest. Platform work itself — infrastructure as code, CI/CD, Kubernetes delivery, observability — is scoped as a bounded project, typically £10,000–£50,000 depending on migration and multi-account complexity.
| Engagement | Price |
|---|---|
| Cloud Cost Audit | From £4,950 |
| Cloud / DevOps project | Typically £10,000–£50,000 |
UK company and contracting entity · UK GDPR-aware delivery · Senior in-house delivery · Code and documentation owned by you
What does infrastructure as code cover?
Infrastructure as code defines cloud resources — networks, compute, databases, permissions — in version-controlled configuration instead of hand-edited consoles, so environments are reproducible, reviewable, and recoverable. Revenant Systems writes infrastructure as code in Terraform, with Terragrunt to keep multi-environment and multi-account estates maintainable.
The founder has built and run Terraform-managed estates on AWS and Azure in prior roles, using plain Terraform and Terragrunt as each estate warranted.
- Terraform and Terragrunt
- AWS, Azure, and Google Cloud
- Multi-environment and multi-account layouts
- Codifying existing hand-built estates
What is GitOps, and why use it for Kubernetes delivery?
GitOps is a deployment model in which the desired state of a Kubernetes cluster lives in Git and an operator such as ArgoCD keeps the cluster matched to it — every change reviewed, versioned, and reversible. Revenant Systems builds GitOps delivery on Kubernetes with Helm charts and ArgoCD.
In prior roles the founder deployed and ran ArgoCD-driven GitOps on EKS, managing production services and a full observability stack through Helm charts and Git.
- Kubernetes (EKS)
- Helm
- ArgoCD
- CI/CD pipelines
Can you reduce our cloud costs?
Usually, yes. Cloud estates that have grown organically tend to carry over-provisioned compute, idle environments, and architecture priced for a scale that never arrived. A cloud cost review applies Revenant's audit-first methodology to spend: measure what the estate actually does, then re-architect the expensive parts with evidence.
In a previous role the founder led a cloud-spend redesign that cut a platform's AWS infrastructure costs by around 60% while its traffic doubled.
- Audit-first spend review
- Right-sizing compute and storage
- Architecture-level savings, not instance tweaks
Observability: metrics, logs, and dashboards
Observability is being able to see what production is doing — metrics, logs, and dashboards that answer questions during an incident, not after it. Revenant Systems builds observability on VictoriaMetrics and Grafana for metrics and dashboards, with Grafana Loki for log aggregation, deployed and managed as code.
Before Revenant, the founder ran VictoriaMetrics, Grafana, and Loki — via Helm and ArgoCD — as the observability stack for platforms processing billions of events a month.
- VictoriaMetrics
- Grafana dashboards and alerting
- Grafana Loki log aggregation
Does every project need Kubernetes and Terraform?
No. Infrastructure should be sized to the team that runs it: a three-engineer startup rarely needs the platform a fifty-engineer organisation does. Revenant Systems recommends Kubernetes, Terraform, or GitOps only where the audit shows they pay for themselves, and works cleanly with whatever infrastructure you already have.
What's included
- Terraform / Terragrunt infrastructure as code (AWS, Azure, Google Cloud)
- Kubernetes deployment and GitOps (EKS, Helm, ArgoCD)
- CI/CD pipelines
- Observability: metrics, logs, dashboards
- Cloud cost review and right-sizing
Stack Terraform · Terragrunt · AWS · Azure · Google Cloud · Kubernetes · Helm · ArgoCD · Grafana
Frequently asked questions
Which cloud platforms do you cover?
AWS, Azure, and Google Cloud, managed through Terraform — with Terragrunt where multi-environment or multi-account estates warrant it. The founder's deepest experience is on AWS and Azure, built up in prior roles; the tooling recommendation follows what you already run rather than a preference.
Our infrastructure was built by hand — can it be brought under code?
Yes — codifying existing hand-built estates is part of the service. The work brings the estate under version-controlled Terraform configuration, so a rebuilt environment matches the one it replaced instead of depending on whoever remembers the console clicks.
How much could we save on our cloud bill?
The audit-first answer: measure before promising. The founder's largest result on this work was around 60% off a platform's AWS bill while its traffic doubled — a prior role, a different estate, and not a figure to expect by default. The fixed-scope cloud cost audit establishes what your own estate can support.
Do you provide monitoring and alerting?
Yes. Metrics and dashboards run on VictoriaMetrics and Grafana, logs aggregate through Loki, and the whole stack is defined in code alongside the estate it watches. Alerting is configured in the same pass, tuned to what genuinely warrants waking someone.
Can you set up our Kubernetes delivery?
Yes. The desired state of the cluster goes into Git, ArgoCD reconciles the running cluster against it, and Helm packages the releases — so a rollback is a revert rather than a rebuild. Handover includes the documentation and access your team needs to run it independently.
We already have infrastructure — can you work with it?
Yes. An existing estate is the normal starting point, and the first job is establishing what you actually have rather than proposing changes to it. Where the current setup does its job, it stays; where it is the source of the problem, the case for replacing it arrives with the evidence behind it.
Every engagement follows the same process — see how we work.
Cloud costs climbing? Let's talk.
Get in touchYour message is read by the engineer who would scope the work; the reply is a short technical conversation.