GCP
On Google Cloud, BuiltForProd delivers the GCP Enterprise Baseline and the GCP blueprints, all production-ready. The Baseline is a production-ready GCP foundation engineered around the Google Cloud Architecture Framework; the blueprints are working workloads on top of it. Each has a public overview and a full documentation set for customers who hold it.
What the GCP offering includes
| Product | Repositories | Overview |
|---|---|---|
| GCP Enterprise Baseline | acme-gcp-platform-baseline | GCP Baseline overview |
| Web App Blueprint | acme-gcp-blueprint-webapp-infra, -code, -gitops | Web App Blueprint overview |
| Data and ETL Blueprint | acme-gcp-blueprint-etl-infra, -code | Data and ETL Blueprint overview |
| Secrets Blueprint | acme-gcp-blueprint-secrets | Secrets Blueprint overview |
| Relational App Blueprint | acme-gcp-blueprint-relapp-infra, -code | Relational App Blueprint overview |
| Multi-tenant SaaS Blueprint | acme-gcp-blueprint-saas-infra, -code | Multi-tenant SaaS Blueprint overview |
| Serverless and Integration Blueprint | acme-gcp-blueprint-serverless-infra, -code | Serverless and Integration Blueprint overview |
| Warehouse and BI Blueprint | acme-gcp-blueprint-analytics-infra, -code | Warehouse and BI Blueprint overview |
| GenAI RAG Assistant Blueprint | acme-gcp-blueprint-rag-infra, -code | GenAI RAG Assistant Blueprint overview |
| Agentic AI Platform Blueprint | acme-gcp-blueprint-agents-infra, -code | Agentic AI Platform Blueprint overview |
| VM Migration Landing Blueprint | acme-gcp-blueprint-vmlanding-infra | VM Migration Landing Blueprint overview |
| Backup and DR Blueprint | acme-gcp-blueprint-backupdr-infra | Backup and DR Blueprint overview |
| FinOps Cost Analytics Blueprint | acme-gcp-blueprint-finops-infra | FinOps Cost Analytics Blueprint overview |
| Observability Blueprint | acme-gcp-blueprint-observability-infra | Observability Blueprint overview |
The acme prefix is the namespace and changes per customer. The blueprint catalog lists every blueprint.
The Baseline on Google Cloud
One Google Cloud organization with 15 projects in three folders: core holds ten single-purpose projects (root, audit, security, identity, network, dns, artifacts, auto, corp, public), backup holds the protected backup project acme-core-backup on its own, and plat holds sandbox, dev, staging and prod. acme-core-network is the Shared VPC host, with a hub VPC and one VPC per isolation domain, prod and nonprod, peered with the hub and never with each other; the plat projects use their domain's subnets, and every VPC egresses through its own Cloud NAT. 19 organization policy constraints and an IAM deny policy hold against project Owners, Security Command Center and 13 log-based detections watch the organization, and two aggregated sinks send every project's audit logs to acme-core-audit. People get IAM roles through 14 Google groups, with Owner only through Privileged Access Manager; pipelines use Workload Identity Federation and impersonate each project's deployer. The architecture overview has the diagram and the project table; the Baseline product page has the deployment options.
The blueprints on Google Cloud
- Web App: a regional GKE Standard cluster with private nodes, Dataplane V2 and node auto-provisioning; the API behind a GKE Gateway with Cloud Armor; the React front end on Cloud Run behind Cloud CDN; Firestore with MongoDB compatibility and Memorystore for Valkey without passwords; External Secrets, external-dns, cert-manager and ArgoCD; a Flask API and a GitOps repository ArgoCD syncs.
- Data and ETL: raw, processed and curated Cloud Storage buckets; a Dataplex Universal Catalog lake that publishes the files as BigQuery tables; BigQuery datasets with policy tags; Eventarc on the raw bucket and an internal Cloud Run service that starts one Dataproc Serverless batch per input file.
- Secrets: SOPS-encrypted files per application per stage, Cloud KMS keys per stage, and a sync to Secret Manager in each stage project.
- Relational App: a Flask API and a React front end on Cloud Run behind a global load balancer with Cloud Armor, on Cloud SQL for PostgreSQL over Private Service Connect.
- Multi-tenant SaaS: a multi-tenant web app on Cloud Run and Cloud SQL for PostgreSQL with a host name per tenant behind a global load balancer with Cloud Armor: row-level security, invite-only Identity Platform sign-in with per-tenant SSO, a control plane on Workflows that onboards and offboards tenants, per-tenant usage metering and dedicated stacks for premium tenants.
- Serverless and Integration: an event-driven orders platform on Cloud Run behind a global external Application Load Balancer with Cloud Armor: Pub/Sub events, a Workflows saga started by Eventarc that compensates on failure, queues with dead letters and replay, and Firestore.
- Warehouse and BI: BigQuery mart datasets built by Dataform from versioned SQL, with assertions, a Looker Studio report and an optional BigQuery reservation.
- GenAI RAG Assistant: a chat assistant that answers from your documents with citations: Gemini on Vertex AI, pgvector on Cloud SQL, Model Armor and Identity-Aware Proxy.
- Agentic AI Platform: an operations assistant on Vertex AI Agent Engine with MCP tools, propose-then-confirm writes, Memory Bank, Model Armor and an evaluation gate.
- VM Migration Landing: migrate to Virtual Machines connected to the landing zone, so each migrated server comes up on the private
vmssubnet with its own identity, Shielded VM, OS Login and no public address; golden images with a sample fleet behind an internal load balancer; VM Manager patching, the Ops Agent, IAP access and backups to the isolated vault from day one. - Backup and DR: Cloud SQL backups in the isolated Backup and DR vault and Firestore backup schedules in every stage, with a daily backup check and a monthly tested restore with alerts.
- FinOps Cost Analytics: a daily cost anomaly alert next to the Baseline budgets, and BigQuery cost views by repository, team, stage and service over the Cloud Billing export, with FOCUS columns, AI cost views and a Looker Studio report.
- Observability: service level objectives with two-window burn-rate alerts, uptime checks of the published URLs, log alerts, Cloud Monitoring dashboards and an OpenTelemetry endpoint for each stage, alerting the stage operations mailbox.
The blueprints page explains how each reads the Baseline through Parameter Manager and how to customize it.
GCP services and tools used
The Baseline and blueprints use Resource Manager, the Organization Policy Service, IAM and IAM deny policies, Cloud Identity, Privileged Access Manager, Workload Identity Federation, service accounts, Security Command Center, Artifact Analysis, Cloud KMS, Secret Manager, Parameter Manager, Google Cloud Armor, Cloud NGFW firewall policies, VPC Service Controls, Cloud Logging, Cloud Monitoring with its SLOs, uptime checks and dashboards, the Telemetry API (OTLP ingestion), Pub/Sub, Cloud Billing with its export to BigQuery, the Backup and DR Service, Workflows, Cloud Scheduler, the BigQuery Data Transfer Service, VPC networks, Shared VPC, VPC Network Peering, Cloud NAT, Private Service Access, Cloud DNS, Identity-Aware Proxy, Compute Engine, Cloud Load Balancing, Cloud CDN, Certificate Manager, the GKE Gateway, Cloud Storage, BigQuery, Dataplex Universal Catalog, Dataproc Serverless, Eventarc, Firestore with MongoDB compatibility, Memorystore for Valkey, Artifact Registry, Cloud Run, Google Kubernetes Engine and Looker Studio. The Multi-tenant SaaS, Serverless and Integration and VM Migration Landing blueprints add Cloud SQL for PostgreSQL over Private Service Connect, Identity Platform, Firestore in Native mode, Migrate to Virtual Machines and VM Manager. Around them: OpenTofu, Terragrunt, GitHub Actions with Workload Identity Federation, SOPS, ArgoCD, Helm, Kubernetes, cert-manager, external-dns, the External Secrets Operator, Checkov, Trivy, tflint, pre-commit and Docker. The GCP services used index links every one to a short explainer of what it does and where the platform uses it.
Tooling versions
| Tool | Pinned version |
|---|---|
| OpenTofu | 1.12.6 |
| Terragrunt | 1.1.6 |
hashicorp/google, hashicorp/google-beta | ~> 8.5 |
| Other providers | hashicorp/helm ~> 3.3, hashicorp/kubernetes ~> 3.2 |
| tflint | v0.64.0, with tflint-ruleset-google 0.40.0 |
No registry module is used: every module lives in its repository. Every GitHub Action is pinned to a commit SHA; the release notes list the versions.
Purchasing
The GCP products are purchased directly from BuiltForProd, set up by the BuiltForProd team and handed over with their repositories and documentation. The steps and the license are on purchasing and licensing.