Build artefacts Target Platform settings

The Build artefacts settings tell DeltaVault where to write your generated bundle, how to run it in Databricks, and which storage locations and schemas to use at runtime. An administrator sets these once at the organization level, and individual projects can override any value for their own deployment.

Where you configure them

Open Settings from the left sidebar, choose Target Platform on the left, and select the Databricks platform card. The fields appear below it under Databricks settings, in three groups: Orchestration, Environment, and Data Loading Behavior. These settings are organization-wide by default: every project in your organization uses them unless a project provides its own override. Editing requires the administrator role.

Orchestration

The Orchestration group controls how Databricks runs the generated bundle when you deploy it.

Job Cluster sets the cluster configuration that Databricks uses when it executes your generated jobs. Specify the cluster policy, node type, and runtime version here so that every job in the generated bundle runs on a consistent, predictable cluster.

Git Source is the Git repository reference that Databricks resolves when deploying your bundle. It tells Databricks where to pull the generated source files from, so the deployed pipeline stays in sync with what DeltaVault committed.

Asset Bundle Path is the root path inside your repository where the generated Databricks Asset Bundle is written. All generated resource files and source notebooks are placed under this path when DeltaVault commits a new bundle.

Environment

The Environment group controls the runtime paths, schemas, and formatting options that the generated notebooks and pipelines use when they execute.

Build Output Path is the folder in your Databricks workspace or storage location where generated build outputs are written at runtime. Notebooks and pipeline stages that produce intermediate outputs write them here.

Repository Name is the name of the Databricks repository registered in your workspace that holds the generated bundle. The generated pipeline definitions reference this name when resolving source file paths.

Notebook Path is the workspace path where the generated notebooks are placed. Databricks resolves this path when it runs jobs that reference individual notebooks from the bundle.

Append Notebook Name controls whether the notebook filename is appended to the notebook path when constructing the full path to each generated notebook. Enable this when your notebooks are stored in a flat directory and need their filename included in the resolved path.

Temp Table Schema is the database schema that the generated pipelines use for temporary tables created during pipeline execution. Isolating temporary tables into their own schema keeps them separate from your permanent output tables.

Table Owner is the principal that the generated SQL assigns as the owner of tables created by the pipeline. Setting an explicit owner ensures that table permissions are consistent across runs, regardless of which user or service account executes the pipeline.

Table Properties are key-value pairs that the generated SQL attaches to every table the pipeline creates. Use this to apply uniform tags, governance labels, or platform-specific properties to all output tables without editing individual notebooks.

Read Files Options are options passed to the file-reading step in your generated pipelines when source files are read from cloud storage. Refer to your Databricks documentation for the supported option keys.

Copy Format Options are format-specific options included in the generated data-loading statements. Use them to control how each file format is interpreted when data is loaded into tables.

Copy Options are additional options added to the generated data-loading statements, controlling load behavior such as error handling. They appear alongside the format options in every generated load statement.

Display Time Zone is the time zone applied to date and time values in the generated output for display and formatting purposes. Generated notebooks reference this setting when formatting timestamps shown to users.

Data Time Zone is the time zone applied to date and time values written to tables by the generated pipelines. Use it to keep the timestamps stored in your output tables consistent.

Data Loading Behavior

The Data Loading Behavior group controls what the combined staging load produces.

Staging Mode sets what that load keeps in the staging table: Delta (only genuine changes) or All (the whole landed batch). Persistent staging always receives only genuine changes, whichever you choose.

Notebook Language sets the language the combined staging load is generated in, SQL or Python. Both choices do the same work in one notebook, the staging load and the persistent staging append. A table bound to a legacy template keeps whatever that template generates.

Overriding per project

Set your organization-wide values in Settings, then open any individual project and go to its Target Platform tab to override specific fields for that project. You only need to fill in the fields that differ from the organization values: any field left blank in the project falls through to the organization setting.

Author platform defaults

Platform administrators can set the Databricks Target Platform values that every organization starts from. Sign in as a platform administrator, open Settings, choose Target Platform, and select the Databricks platform card. At this platform scope the toggles and the Orchestration and Environment fields are editable, and the values you save become the shared baseline for Databricks. Only the Databricks platform is editable here today; Snowflake and Fabric show a read-only message.

Organizations inherit the platform baseline in three ways:

Live while unset. An organization that has not set any Databricks value of its own reads the platform baseline directly, and that value updates in place when a platform administrator changes it.
On platform select. When an organization first selects Databricks as its target platform, its Databricks settings are seeded from the platform baseline, so it starts from the shared default instead of a blank slate.
Apply platform defaults. The organization-facing Apply platform defaults action writes the team-authored baseline into the organization’s Databricks settings, not just the built-in factory values.

Once an organization changes any Databricks value, it owns that configuration and stops inheriting later platform edits. Editing at the platform scope requires the platform administrator role.