GCP reference architectureApache Superset on GCP
Deploy a fully-managed Apache Superset instance in your GCP project on GKE Autopilot, backed by Cloud SQL and Memorystore. D23 integrates your dbt semantic layer, connects BigQuery natively, and manages row-level security and Google SSO.
GCPRuntime: Google GKE + Cloud SQL + MemorystoreEntry point
One-click deploy from the D23 console or Terraform module
Cloud / GCP
D23 platform
Apache Superset
dashboards · datasets · SQL Lab
D23 Managed Layer
SSO · row-level security · upgrades
Compute & cache
Kubernetes cluster
Google GKE + Cloud SQL + Memorystore
Data Modelling AI
managed by D23
BigQuery Dataset Mapper
managed by D23
dbt Semantic Layer Sync
managed by D23
Async Query Worker
managed by D23
Data sources
50+ supported databases
Your infrastructure
Google GKE Autopilot
Cloud SQL PostgreSQL (metadata)
Memorystore Redis
BigQuery
How the deployment works
Starting point: One-click deploy from the D23 console or Terraform module
- 1
D23 provisions a GKE Autopilot cluster in your GCP project and deploys Apache Superset via Helm
- 2
Cloud SQL (PostgreSQL) is provisioned as the Superset metadata store; Memorystore for Redis backs Celery workers and query caching
- 3
Workload Identity and Private Service Connect are configured so Superset queries BigQuery and Cloud SQL without public exposure
- 4
D23 AI connects to BigQuery, runs schema discovery, and models your dbt semantic layer into Superset datasets
- 5
Ongoing: D23 manages zero-downtime Superset upgrades, row-level security, Google SSO (Cloud Identity), and scheduled async queries
D23 managed services
- Data Modelling AI
- BigQuery Dataset Mapper
- dbt Semantic Layer Sync
- Async Query Worker
Connected data sources
50+ databases supported — Snowflake, BigQuery, Redshift, Postgres, ClickHouse, and more.
Infrastructure & data
Google GKE + Cloud SQL + Memorystore
Google GKE Autopilot
Cloud SQL PostgreSQL (metadata)
Memorystore Redis
BigQuery
Security & operational considerations
- Workload Identity eliminates static service-account keys entirely
- Private Service Connect ensures BigQuery traffic stays on Google's backbone
- Cloud Audit Logs capture every query and admin action
- dbt semantic layer is version-controlled and applied automatically on upgrade
GCP deployment
Deployed in your GCP project on GKE Autopilot, with Workload Identity, Private Service Connect to BigQuery, and Cloud Audit Logs. D23 manages the dbt semantic layer and zero-downtime upgrades.
Turn every team into a data team.
Get managed Apache Superset, the dashboards your business needs, and AI on top, without hiring a data team.