Cloud Architecture & Design
Decisions that are cheap now and expensive to reverse later.
We review how your systems fit together, what happens when a component fails, where the data lives and what a regional outage would do to the business. Then we write the target architecture with the trade-offs of each option stated, so the choice is yours and the reasoning is on paper.
Tools we work in
- AWS
- Azure
- Google Cloud
- Terraform
- Kubernetes
- Cloudflare
- PostgreSQL
Scope
What this covers
The specific work included when you engage us for this service.
- A review of the current architecture against reliability, security, cost and operability
- Failure analysis covering what breaks, what degrades and what fails sideways
- A target architecture with the trade-off of each option written down
- Network design, private connectivity, DNS and load balancing
- An environment strategy for development, staging and production
- A staged path from the current design to the target one
What you receive
- 01A written architecture review naming the risks in the current design and their consequences
- 02A target architecture document with diagrams, decisions and the alternatives considered
- 03A staged implementation plan with cost and effort attached to each stage
Questions
Cloud Architecture & Design, answered
We already have architects. What does this add?
A second opinion with production scars, usually. Internal architecture decisions are made by people who will have to live with them, which is healthy, but it also makes it harder to argue against the option the most senior person prefers. We are not invested in the outcome, so we can say that a design will be expensive to operate or impossible to recover, and put the reasoning in writing where it can be debated.
How do you decide between a monolith and microservices?
By team size and deployment pressure rather than by fashion. A modular monolith is the correct default for most organisations, because a single deployable unit is far cheaper to operate. Services earn their cost when separate teams need to release independently or when one component has genuinely different scaling needs. Splitting early usually produces the operational burden of a distributed system without the organisational benefit.
Do you produce a diagram, or just a document?
Both, and the diagram is expected to be maintainable rather than beautiful. We use text-based diagram sources where the tooling allows, so the architecture lives in the repository next to the infrastructure code and changes with it. A diagram that only exists in a slide deck is out of date within a quarter.
Most often requested by
Related
More in Cloud, DevOps & Security
Cloud Migration
Move to the cloud without carrying the old problems with you.
Cloud Cost Optimisation
Most cloud bills contain spend that nobody decided to make.
CI/CD & Release Engineering
Deploys that are boring, which is exactly the point.
Kubernetes & Container Platforms
Containers are quick to adopt and slow to master.
Tell us where the manual work is
A 30-minute call. We map the process, tell you what we would automate or build first, and what it costs. No retainer before you agree the scope.
- No retainer before scope is agreed
- Written plan you can act on alone
- Worldwide delivery, on your hours