Topic directory
Use topic hubs to navigate the publication like a working technical library rather than a flat archive.
How this is organized
Most AWS content is organized by service: "here's an S3 tutorial, here's a Lambda guide." That's fine for learning individual services. It's useless when you're trying to make an architecture decision that spans six services and three teams.
InfraTales organizes by the problem you're solving. If you're designing a system shape, that's architecture. If you're figuring out why your deploy pipeline keeps breaking, that's DevOps. If you're trying to cut your AWS bill without breaking production, that's cost optimization. The service names show up inside the articles - they're not the organizing principle.
Each topic hub below collects articles, patterns, and case studies around one operating surface. Some have deep libraries already. Others are building out. Here's what each one covers.
Where you go when the question is "how should this system be shaped?" VPC design, multi-account strategy, service boundaries, data flow patterns. Not service tutorials - architecture decisions with trade-offs, failure modes, and cost implications baked in.
If you're doing a design review or evaluating whether your current architecture will hold at 10x scale, start here.
CI/CD pipelines, IaC patterns, developer platforms, deployment strategies. The mechanics that let teams ship without breaking things. CDK vs Terraform decisions, blue-green vs canary, and what your deploy pipeline should actually look like when it's not a demo.
This is where you'll find the CodeDeploy, CodePipeline, and GitHub Actions content.
IAM that actually follows least-privilege (not just claims to), failure isolation patterns, KMS encryption decisions, WAF rules that do something useful, and recovery plans that work at 2am. The stuff that matters when something breaks and the on-call engineer is you.
If your security review is coming up or you just had an incident, this is your starting point.
Not just "turn off unused instances." Architecture-level cost decisions, FinOps patterns, Reserved Instance math that doesn't lie, and the specific places where optimization breaks production reliability. Real numbers, real trade-offs.
Every cost article includes a breakdown by service so you can compare against your own bill.
Production AI isn't a notebook running on a GPU. It's RAG pipelines, vector stores, model serving, LLM cost management, and the operational reality around inference workloads. Bedrock, SageMaker, OpenSearch Serverless - the infrastructure side that most AI content ignores.
Content building out - first articles planned for Q2 2026.
Monitoring that catches problems before users do. Logging that's actually searchable during an incident. Alerting that doesn't wake you up for things that can wait. CloudWatch, X-Ray, OpenTelemetry, and the operational discipline that makes systems legible under stress.
Content building out - first articles planned for Q2 2026.
Architecture decisions, service boundaries, and production patterns for building on AWS.
Delivery systemsDeveloper platforms, deployment workflows, infrastructure as code, and the mechanics that keep teams shipping.
AI infrastructure — coming soonProduction AI infrastructure, model-serving trade-offs, vector systems, and the operational realities around LLM workloads.
ResilienceIAM, failure isolation, recovery planning, defensive controls, and reliability work that survives the first incident.
Visibility — coming soonMonitoring, logging, alerting, on-call ergonomics, and how to make systems legible under stress.
EfficiencyFinOps, savings trade-offs, architecture-driven spend reduction, and the places optimization breaks production.
Evidence — coming soonRetrospectives, delivery narratives, and real-world evidence of how decisions behave outside clean-room examples.
Judgment — coming soonStrong opinions, comparison pieces, and architecture review notes for teams choosing between real trade-offs.
How the taxonomy works
Every article should belong to one primary hub and a small set of public facets. Internal tags stay internal and carry format, funnel, series, and refresh metadata.
AWS Architecture, DevOps and Platform Engineering, AI and GenAI Infrastructure, Security and Reliability, Observability and Operations, Cost Optimization, Case Studies and Failure Breakdowns, Opinion and Architecture Reviews.
Use narrowly and consistently: serverless, kubernetes, terraform, cdk, networking, iam, finops, genai, observability, and multi-region.
Reserve internal tags for content type, series, lifecycle, and intent so the public taxonomy stays readable and scalable.