Inference Public

AWS foundations and landing zones

Establish the foundations.

AWS environments with clear ownership, controlled access and room to grow.

Cloud engineers reviewing an account and network structure on a wall screen.

Foundations someone can own

An AWS estate grows one account, one network and one exception at a time. Without agreed structure, nobody can say who owns an account, why a permission exists or which team pays for shared networking, and every new service waits on a ticket.

We design and build foundations that fit how your organisation actually works: account boundaries that follow ownership and policy, identity that federates with your existing directory, connectivity with clear routing ownership, and controls that are enforced rather than documented.

When to bring us in

  • You are setting up AWS for a new service or department and want the structure right before teams arrive.
  • You have inherited accounts, permissions and networks that nobody fully understands.
  • Teams wait days for accounts, network changes or access, or work around the controls to avoid waiting.

What we take on, and what stays with your team

We take on: the design, the infrastructure as code, the provisioning and onboarding procedures, and the evidence that boundaries and controls behave as intended.

Your team keeps: decisions about who owns which accounts, which risks are acceptable, which exceptions are granted, and the day-to-day operation once handed over.

Engineering scope

Accounts and organisational units

Group accounts by the policies they need, not by the organisation chart. Define account lifecycle: request, provisioning, ownership, review and closure. Control Tower orchestrates a governed multi-account environment and Account Factory standardises provisioning; we decide with you where it fits and where it does not.

Identity and privileged access

Federate IAM Identity Center with your identity provider, define permission sets by role, separate day-to-day from privileged access, and keep emergency (break-glass) access tested and monitored.

Logging, detection and controls

Organisation-wide CloudTrail to a protected log archive account, AWS Config, GuardDuty and Security Hub with delegated administration, and service control policies for the guardrails that must never be crossed.

Connectivity and DNS

Transit Gateway with route tables that segment environments and teams, Site-to-Site VPN or Direct Connect for hybrid access, Route 53 Resolver for hybrid DNS, and IP address planning that survives growth. Central networking is owned by one team and consumed by application teams through a defined request contract.

Drift and exceptions

Detect drift from the intended configuration and keep a register of exceptions, each with an owner, a reason and an expiry date.

An illustrative project

Illustrative example, not a client engagement

Starting state. A department runs a dozen AWS accounts created by different projects. Access is granted by individual IAM users, networking is peered ad hoc, and audit logging is inconsistent.

Agreed work. Agree an OU structure and account ownership, introduce federated access with role-based permission sets, centralise logging, and replace peering with a Transit Gateway design for new workloads.

Deliverables. Infrastructure as code for the foundation, an account provisioning procedure, an access model, a connectivity design with routing ownership, an exceptions register and runbooks.

Acceptance. An agreed new account is provisioned through the procedure, access boundaries are demonstrated (who can and cannot do what), connectivity works as designed, and the operating team performs an onboarding themselves using the delivered procedure.

What we need from you

  • An owner for the foundation, and the people who decide security policy and network standards.
  • Access to the management account or a delegated administrator, and to the identity provider configuration.
  • Your existing standards, constraints and any assurance requirements the foundation must evidence.

Costs, risks and trade-offs

Centralised controls reduce risk but can slow teams down; we make that trade-off explicit and give teams a self-service route for common requests. Transit Gateway simplifies routing at scale but has hourly and data processing charges that peering does not.

Shared networking, logging and security tooling need an agreed allocation, or they become the part of the bill nobody owns.

FinOps and cloud cost engineering covers how running costs are made attributable and managed.

Related

Read A landing zone needs owners as well as accounts and The trouble with VPC peering at scale.

See anonymised experience. Compare ways to work together.

A sensible first step

Bring a description of the estate as it is, the services coming next and what worries you most. We will discuss whether a foundations review, a build or an embedded engineer is the right starting point.

A first conversation does not commit either side to anything, and we will not ask for access to your systems until a scope is agreed.

Discuss AWS foundations