Production-grade engineering reference for automating cloud billing, cost governance, and FinOps workflows across AWS, GCP, and Azure.

🌐 www.cloud-cost-optimization.org

Cloud spend has outgrown dashboards. At enterprise scale, FinOps is an engineering discipline: deterministic pipelines that ingest billing telemetry, normalization layers that converge divergent provider schemas into a single dimensional model, and governance controls that enforce tagging, budgets, and anomaly detection before raw cost data ever reaches finance. This site is the reference for the engineers who build and operate that pipeline — architecture first, runnable Python second, operational failure modes third.

Every page is written for FinOps engineers, cloud architects, DevOps, and Python automation builders who read code before prose. No vendor pitches, no generic advice — just reference architectures, complete and runnable implementations, schema tables, and the failure modes you actually hit in production.

The material is organized into four streams, each a deep, interlinked set of guides:

• FinOps Architecture & Billing Fundamentals — the data flow that turns raw cloud telemetry into trustworthy cost intelligence: Cost Explorer architecture, BigQuery billing export, Azure Cost Management, reserved-instance mapping, cross-cloud allocation, commitment-discount optimization, and the FinOps operating model.
• Cloud Billing Data Ingestion & Parsing — provider-specific extraction, retries, and state tracking: the AWS CUR-to-data-lake pipeline, GCP BigQuery export sync, the Azure Cost Management API, rate-limit and retry handling, async processing, Pub/Sub streaming, and multi-cloud schema normalization.
• Resource Tagging & Validation Pipelines — the metadata contracts allocation, showback, and lifecycle governance depend on: schema-driven tag validation, AWS Config enforcement, shift-left policy checks in Terraform, and automated, idempotent remediation.
• Cloud Cost Anomaly Detection & Budget Automation — turning cost signals into statistically-grounded alerts and budget guardrails: rolling z-score and seasonal-decomposition models, budget-alert webhook handlers, alert routing, and CI/CD cost guardrails that block over-budget changes before they merge.

• Runnable, import-complete Python. Every code block is real: boto3, google-cloud-*, azure-mgmt-*, pandas, retry decorators, structured logging, dataclass models, and idempotency keys — not pseudocode.
• Architecture-first. Each topic opens with the constraint that actually breaks naive approaches at scale — finalization lag, rate limits, schema drift, eventual consistency — then shows the pattern that survives it.
• Deterministic by default. Schema contracts, idempotent writes, content-addressed keys, partition-aware queries, and observable feedback loops throughout.
• Densely cross-linked. Guides reference each other the way a real system's components do, so you can follow a concept from a reference architecture down to a focused how-to and back.

• CUR vs Cost Explorer API, BigQuery export vs Pub/Sub streaming, and Azure EA vs MCA billing scopes — the decision guides for the choices that shape a billing pipeline.
• Enforcing tag policies in Terraform plan checks, GitHub Actions cost-diff reporting, and blocking Terraform plans that exceed a budget — cost governance wired directly into CI/CD.
• Detecting cost spikes with a rolling z-score and seasonal decomposition for cloud cost baselines — anomaly detection that respects finalization lag and weekly seasonality.

Built as a static site with Eleventy (11ty), hand-authored inline SVG diagrams, and structured data (JSON-LD: Article, BreadcrumbList, HowTo, FAQPage, Dataset). Deployed on Cloudflare Pages.

Spot an error, have a pattern to add, or want a topic covered? Open an issue — corrections and real-world production war stories are welcome.

Explore the full reference at www.cloud-cost-optimization.org.