What should the system boundaries look like?
Use this hub when you are deciding between service shapes, asynchronous boundaries, state placement, or multi-account and multi-region strategy.
Topic hub
AWS Architecture
Architecture decisions, service boundaries, and production patterns for building serious systems on AWS.
Questions this hub should help answer
Use the topic like a decision surface, not a tag archive.
Which AWS choices survive production?
Look here for articles that go beyond “service A vs service B” and into blast radius, failure isolation, and operational cost.
When should you switch to patterns?
If the decision is still architectural rather than implementation-level, move into the architecture library and then come back for deeper reads.
Browse architecturesStart here in this topic
The strongest first read in this area.
Open one article that gives the clearest view of how this problem space behaves in production, then continue into the wider set below.
AWS CDK Transit Gateway Multi-Environment VPC: Dev, Staging, Prod on One TGW
Running dev, staging, and prod on separate VPCs sounds clean until you need them to talk — and your Aurora failover has never actually been tested. This post walks through a real CDK TypeScript stack wiring Transit Gateway, Aurora Global Database, ECS Fargate, and Global Accelerator across three env
Within this topic
Then move through the rest of the hub.
These pieces stay inside the same operating surface and are better for depth once you already have the context from the spotlight read.
AWS CDK Web App Deployment: VPC, Fargate, Aurora, and CloudFront Wired Together
Most CDK tutorials stop at 'it deployed.' This one covers the full aws cdk web application deployment stack - VPC isolation, Fargate on private subnets, Aurora with KMS, CloudFront, Secrets Manager rotation - and the six operational gaps that will wake you up at 2am if you skip them.
Continue from here
Move to the adjacent surface when the problem broadens.
Read the wider publication
Go back to Start Here if you want the best cross-topic entry points rather than staying inside a single hub.
Open Start HereNeed buyer-side proof?
Case studies and failure breakdowns are where the publication shows how decisions behave under delivery and production pressure.
Read case studiesNeed direct help?
Consulting is for architecture reviews, cost teardowns, and AI infrastructure assessments that need direct judgment instead of more reading.
View consulting