Organization, folders, and project boundaries
Translating enterprise environment patterns into GCP without losing clarity around ownership and separation.
Lab
Ongoing work around GCP identity, hierarchy, BigLake governance, storage design, and how enterprise platform standards carry cleanly into a second cloud.
Focus areas
Translating enterprise environment patterns into GCP without losing clarity around ownership and separation.
Service accounts, remote state, Workload Identity Federation, and least-privilege bootstrap design.
Testing where governed storage access should remain external and where warehouse materialization adds value.
Applying governed access patterns directly inside the GCP build while the platform design continues to mature.
Execution lanes
The GCP build is not one vague track. It is moving through a few deliberate workstreams with clear platform responsibilities.
Focus repos: terraform-foundation-gcp and platform-docs-gcp
Building a clean project-per-environment shape so platform boundaries stay explicit from the start.
Focus repos: terraform-bootstrap-gcp and internal auth docs
Treating bootstrap identity and least-privilege access as platform architecture, not setup trivia.
Current emphasis: Silver storage, Gold serving, and policy boundaries
Defining where external governed access should remain in storage and where warehouse-native tables should take over.
Current decisions
These are active architecture decisions inside the build, not abstract prompts. Each one affects how the platform gets separated, governed, and made operable across environments.
One project per environment, reusable logic, and no quiet hardcoded drift across dev, staging, and prod.
Not just browser login, but also CLI auth, ADC, impersonation, and federated CI/CD access.
Choosing between object-storage access and native warehouse tables based on governance, cost, and consumer needs.
Engineering discipline
If the setup belongs in Terraform, the lab should treat it that way from the start.
The useful artifact is not only the working state, but also the explanation behind it.
The same logic should deploy across environments through controlled inputs, not ad hoc edits.
The documentation should stay accessible and precise while remaining faithful to enterprise practice.