Cloud & infrastructure consulting and management

The cloud isn’t an end in itself, it’s a tool: the challenge is choosing the right service, sizing it correctly and keeping it under control over time. We design and manage cloud infrastructure in a provider-agnostic way, so the solution follows your needs rather than a vendor’s price list. No black boxes: documented configurations, transparent costs and a single technical point of contact who owns the whole stack.

Overview

We support you across the full infrastructure lifecycle: from initial design to migration, from day-to-day management to cost optimisation. We work mainly on Google Cloud and, alternatively, on AWS and DigitalOcean; on request also Azure, Hetzner and OVH. For each workload we choose the simplest combination that genuinely holds up: a well-sized virtual machine when that’s enough, load balancers and auto-scaling when traffic peaks demand it.

The goal is infrastructure that is reliable, understandable and reproducible - something you can grasp and maintain even without us. That’s why every decision ends up written down: architecture diagram, rationale, expected costs and the procedure to rebuild everything from scratch. If you ever want to switch provider - or consultant - the keys stay in your hands, and whoever comes after finds order, not archaeology.

Typical problems we solve

• The cloud bill keeps climbing every month for no clear reason → cost audit, rightsizing and a report with the savings ranked by priority.
• Everything runs on a single machine that no longer handles the peaks → we redesign with a load balancer, auto-scaling and a managed database.
• Only the person who built it understands the infrastructure → we map it, document it and hand back updated access and diagrams.
• You need to migrate from on-premise or switch provider without stopping the service → inventory, dress rehearsal on a staging environment and a rollback plan.
• Backups exist but no one has ever restored them → we encrypt them, verify them with periodic restores and write the recovery procedure.
• Every service is exposed on a public IP → we segment the network with VPCs, private subnets, NAT and tuned firewall rules.
• The environment grew by hand and no one can reproduce it → where the complexity justifies it we move it to Terraform, otherwise to direct, documented configurations.

What’s included

• Design: service architecture, sizing, provider selection and cost estimates before you spend a euro.
• Migration: moving from on-premise or another cloud, with an inventory, an agreed window and a rollback plan ready to go.
• Operations: updates, monitoring, security patching and on-call cover on a monthly retainer when downtime genuinely costs you.
• Cost optimisation: rightsizing, reserved resources, tiered storage, automatic shutdowns and clear spend reports.
• Infrastructure as code, when it helps: where complexity justifies it we describe everything in Terraform, versioned and reproducible; on simple environments we stay with direct, documented configurations.
• Networking and security: VPCs, subnets, security groups, secrets management, encryption and an attack surface kept to a minimum.
• Backups and high availability: encrypted backups that are genuinely tested, redundancy and restore plans we know how to run.

Stack and technologies

On Google Cloud we use the proven building blocks: Compute Engine, Cloud SQL, Cloud Storage, load balancers and auto-scaling groups; on AWS and DigitalOcean the equivalents, chosen with the same criteria. We manage machines with Debian and Ubuntu Server - on request AlmaLinux, Rocky Linux or others - and design networks with VPCs, public and private subnets, NAT and tuned firewalls. Where complexity justifies it we automate with Terraform, so every resource is versioned code; elsewhere we prefer direct configurations that anyone can read. Rounding out the picture: managed databases, object storage, queues and caches, monitoring and alerting with Prometheus and Grafana or the provider’s native tooling, and repeatable deployment pipelines. If you already run containers or Kubernetes, we plug into your workflow instead of reinventing it.

A real-world example

A company running an internal management system and an e-commerce store had grown “by hand” for years: about a dozen oversized virtual machines, no documentation and a monthly bill on a steady climb. We ran the inventory, measured actual CPU, memory and disk usage over a few weeks, and redesigned the environment on Google Cloud: fewer machines but sized on real data, a managed database service, tiered storage for archives and automatic shutdown of staging environments overnight and on weekends. The migration happened in two agreed windows, with a rollback plan ready and never used. The result: monthly spend cut by roughly a quarter, a document describing every piece of the infrastructure, and backups that now get a trial restore every month.

Who it’s for

• SMEs and professional firms who want a well-managed cloud without standing up an in-house IT department.
• Web agencies and software houses that need solid infrastructure behind their clients’ projects.
• E-commerce and SaaS that need to scale during peaks and keep costs under control the rest of the year.
• Teams already on the cloud who are paying too much, lack documentation or want out of a lock-in.

How we work

We start from an audit of what exists or from a design on paper, define the simplest solution that carries the load, and put it into production with documented configurations - described as code where complexity justifies it. Every maintenance window is agreed in advance, every change tested before touching production. Standard coverage runs Monday to Friday, 10:00-18:00 , with priorities and timeframes described on the response times page; for ongoing management many clients use hour packages, which never expire. Quotes are free, with a reply within one business day: you always know what’s running, what it costs and why.