We design, migrate and operate AWS for SaaS and enterprise teams. You get a named engineer in your Slack channel, not a support portal and a case number.
Sample replay of a real class of incident. It matches the targets we hold ourselves to: page in 2 minutes, a human on it inside 15.
None of them are exotic. They are what happens when infrastructure grows faster than the team that owns it.
Idle RDS instances left over from a load test. gp2 volumes nobody migrated to gp3. NAT Gateway charges from pods chatting across availability zones. On one EKS estate we cut the monthly run rate in half without turning off a single feature.
CloudWatch alarms fire into a Slack channel that everyone muted in month two. The first person who notices is a customer. We take the pager, write the runbooks, and tell you the next morning what changed and why.
Multi-tenant SaaS built on one shared pool works until a single customer runs a bad query at scale. Tenant isolation, per-tenant limits and noisy-neighbour protection are cheaper to build before that call comes in than after.
Most engagements start with one of the first two and grow into the rest.
Your workload measured against the six AWS pillars, with the findings ranked by what will hurt first.
Data centre, other cloud, or a messy half-migration someone else started. Cutover plan, rollback plan, dry run.
Single-tenant product to a real multi-tenant platform: tenant isolation, metering, onboarding that does not need an engineer.
Pool, silo or bridge models on AWS, with per-tenant limits so one customer cannot spend everyone else's capacity.
Lambda, API Gateway, EventBridge, Step Functions. Built where they pay off, avoided where a container is simply cheaper.
Terraform, pipelines, environments that rebuild from code. Deploys that a new hire can run on their first week.
Containers on EKS or ECS, microservices where they earn their keep, and the observability to debug them at 3am.
IAM cleanup, encryption, audit trails and evidence collection for SOC 2 and ISO 27001 questionnaires.
In that order. We do not take the pager for a system we have not read first.
We read your account, not your slide deck. Cost breakdown, IAM exposure, single points of failure, and what breaks first at 3x traffic.
The ranked list gets worked top down. Infrastructure moves into Terraform as we touch it, so the next change is repeatable.
We take the pager. Alarms, runbooks, patching, backups and restore drills that actually get tested.
Reserved capacity, right-sizing, storage tiers and a written report on where the money went and what we removed.
Confidential, financial services
A team running containers on hand-built EC2 nodes needed Kubernetes without inheriting a platform-engineering headcount. We built the cluster in Terraform, moved the workloads across in stages, and handed back deploys their own developers run.
“We have worked with iNube for seven years across a dozen companies. They rebuilt our platform on a microservice architecture using AWS EKS and have kept it running ever since.”
Jaime VillagomezCEO, Tesora Group · tesorainternational.com“The team moved quickly, asked the right questions about our data, and set up infrastructure that scaled with us instead of holding us back.”
Zahra BiabaniCEO, IntheLoop AI · intheloopai.comMonitoring and alarms, 24/7 on-call response, patching, backup and restore testing, security review, cost reporting, and infrastructure changes as your product changes. Application code is a separate engagement, though the same engineers usually do both.
Production alarms page an engineer within two minutes and a human is working the incident inside fifteen. You get a written root cause the next working day, not a ticket status.
No. Most engagements start with a Well-Architected Review and one workload. If the first cutover goes badly you have lost two weeks, not a quarter.
Managed retainers are priced on the number of environments and the response target, not on a percentage of your AWS bill. That way we are not paid more when your spend goes up. Send us your current bill and we will quote against it.
The engineer doing the work. We are a small senior team, so there is no account manager relaying messages between you and the person touching the console.
That is the normal case. We own the infrastructure and the pipelines, your developers keep shipping, and we hand over anything you want to run yourselves.
Thirty minutes with an engineer, no slide deck. We will tell you what we would change first and what it would cost to have us do it.
