Reference architecture
Built for Azure first, with reference designs for AWS and Google Cloud and variants for Microsoft Fabric, Snowflake, and BigQuery. It includes an access model for admins, engineers, and analysts.
A lakehouse defined in Terraform, with access rules and quality checks from the first deploy.
Download the PDFA short PDF to share with your team. It covers what is in the package, how it works, and a worked example.
Your download has started.
We also emailed the link to . It works for 7 days.
Download didn't start? Get the PDF
Want to try it on your own data? Talk to us.
Networking, storage, identity, access policies, and deployment pipelines all have to be right before anyone ships a data product. Teams rebuild them for every program, and the security settings get decided last.
Built for Azure first, with reference designs for AWS and Google Cloud and variants for Microsoft Fabric, Snowflake, and BigQuery. It includes an access model for admins, engineers, and analysts.
Network, storage, Key Vault, Databricks, Unity Catalog, and monitoring. Dev and prod have separate settings, each with its own remote state.
A Databricks Asset Bundle with templates for files through Auto Loader, incremental database loads over JDBC with a watermark, and paginated REST APIs. Data lands in bronze and is typed and tested on its way to silver and gold.
Quality rules in YAML send bad rows to a quarantine table and stop a load that crosses a set threshold. Production deploys log in with OIDC and wait for an approval.
A virtual network with private endpoints, delegated Databricks subnets, and private DNS zones.
ADLS Gen2 with one container per layer: landing, bronze, silver, and gold. Shared keys and public access are switched off.
Key Vault with role-based access only, network deny by default, and purge protection in production.
A Premium Databricks workspace with VNet injection and no public IPs. Production gets a private front end.
Unity Catalog grants. Admins get everything, engineers read and write, and analysts read silver and gold.
Storage, Key Vault, and Databricks audit logs go to Log Analytics.
Guardrail checks run on every pull request and need no cloud access, so problems show up while the change is still a diff.
A change that loosens a security setting fails before it reaches a subscription. Production deployments still wait for a person to approve them.
Ask about the Data Platform Foundation
Dex answers from our own pages and links to where it found the answer.
Show us your landing zone and network layout. We will show you where the modules plug in.