Managed cloud and FinOps: 24x7 operations, named TAM, monthly cost review

What is included

The default scope below is the starting point. Anything not listed is out of scope until it is written into the statement of work.

SLA: what is public and what is contractual

This page does not publish minute counts or uptime percentages. A marketing page that invents response minutes creates a liability. The contract states:

• ✓Severity definitions, and the clock that starts when an alert fires
• ✓Response targets by severity
• ✓The on-call rota and escalation path

AWS, Azure and Google Cloud publish their own service SLAs. Our commitments cover our response and our operations work, not the provider's service levels. When a provider-side event is the cause, we work the provider's support path with you. We do not claim control over it.

The TAM and the three time zones

A TAM is a named person who knows your account. They join architecture reviews, run the monthly cost review, and act as the escalation point once operations has acknowledged an incident. The TAM is not your whole on-call rota. If the TAM is away, the runbooks and the rota must still work. That is how we test whether the service depends on a person or on a practice. A named TAM from us does not depend on an AWS Enterprise Support subscription.

Engineering and operations run from Dhaka. The Korea office and the US office take client contact in their own hours and pass technical requests to Dhaka:

Alerts for in-scope infrastructure are covered around the clock. The overlap between your working hours and our team's is agreed per engagement, and the escalation path is written into the contract.

FinOps: a monthly review, not a promise

Most surprise bills come from idle capacity, unattached storage, forgotten environments, and traffic that pays full origin price. The monthly review names those items while they are still small.

Savings are ranges and should be treated as estimates:

• ✓Reseller pricing on AWS compute and storage has typically come in 15 to 35 percent below list. Where that arrangement applies, the range is written into the proposal.
• ✓CloudFront bandwidth can be up to 40 percent lower under the same arrangement.
• ✓Results depend on the estate. A clean estate saves less; a neglected one may save more.

Three practices run through the review:

• Right-sizing. Match instance, storage and database size to actual use. Idle and orphaned resources are stopped or removed.

• Commitments. Savings Plans and reserved capacity are modelled, then bought only when coverage and cash flow justify them. They are never automatic.

• Governance. Tags, budgets and anomaly alerts, so a new environment cannot hide for a quarter.

A FinOps-only audit is available if you are not ready for operations. It has a fixed scope, and ongoing governance is quoted separately.

Security evidence and engineering on top

Day-to-day controls: least-privilege access, encryption in transit and at rest where each service supports it, secrets handling under a runbook, logs an auditor can follow, and vulnerability scanning of in-scope hosts. The security and trust page has the certificate details. Certification decisions stay with your auditors.

Kubernetes (EKS, AKS or GKE, depending on the estate), CI/CD pipelines, Infrastructure as Code with Terraform or CloudFormation, and network design are scoped as project work. They then hand over to the same operations team, so the platform that was built is the platform that is watched.

How it is priced

• ✓Cloud usage: billed under the arrangement in your contract. Engagements are quoted in USD. Other currencies are available only by contract, so ask.
• ✓Managed fee: monthly, quoted after we have seen the estate or a recent bill.
• ✓FinOps-only audit: fixed scope.
• ✓Project work: quoted separately, then handed to operations.

Taxes apply as applicable in your jurisdiction and the contract.