Azure
Under evaluation. We'll list it here when we have a production system and a team that has run one at 3am.
Plenty of consultancies list every logo on their homepage. We list one. Knowing a single platform down to its failure modes, quotas and pricing quirks is worth more to you than a shallow acquaintance with three.
Every system we've shipped runs on AWS. Landing zones and multi-account structure, serverless and containers, data and analytics, observability, security guardrails and the FinOps work that keeps the bill defensible.
We're deliberately transparent about this: if your organisation has already standardised on another cloud, we are probably not your best option for a greenfield build there — and we'd rather tell you that now.
What travels well is the discipline, not the service names. Infrastructure as code, least privilege, immutable deploys, SLOs, cost attribution — those are platform-independent, and we bring them wherever we work. We also keep the portable layer portable: containers, open standards and Terraform, so nothing about our approach welds you shut.
Ask us the awkward question →The infrastructure layer, versioned, reviewed and reproducible. No console clicking.
Where workloads genuinely need it — and honest advice about when they don't.
Vendor-neutral instrumentation, so your telemetry isn't hostage to a billing decision.
Standard images and runtimes that will move if you ever need them to.
We add a platform when we've shipped production systems on it — not when a client asks whether we could.
Under evaluation. We'll list it here when we have a production system and a team that has run one at 3am.
Under evaluation, particularly for data and ML workloads. Same bar applies.
We use edge compute and CDNs today as part of AWS architectures; a standalone practice would need more mileage.
Bring us your architecture diagram, your bill, or the incident you never want to repeat.