Project Configuration
A project holds one semantic layer: its models, metrics, dashboards, and the environments those resolve against. Project configuration controls how that layer is versioned, which environment is the default, and who can change what.
Most day-to-day modeling happens in Studio; this page covers the settings around it.
Project Basics
Name and Description
Give the project a name that reflects the domain it models — revenue-analytics, product-usage — and a description explaining what belongs in it. Teams that split projects by domain rather than by team end up with far less duplicated metric logic.
Environments
Each project can hold several environments — typically development and production — and each environment binds to one connector. Set the default environment members land in, and mark production so that changes there require review. See Add Environment & Connector for setup.
Defaults for Metrics
Set the project-level default time dimension, grain, and reporting window that new metrics and explorations inherit. Individual metrics can override any of these in their definition.
Versioning and Deployment
Model Versions
Every save in Studio creates a version of the model. The version history shows what changed, who changed it, and when, and any version can be restored. Metric definitions are the contract your dashboards depend on, so treat a restore the same way you would treat a revert in source control.
Promote Between Environments
Promote a validated model version from development to production. Aptrics compares the two, lists the metrics and dimensions that would change, and flags any dashboard tile that depends on something being removed or renamed before you confirm.
Export and Import Definitions
Models can be exported as OSI-specification YAML and checked into your own repository, then imported back into another project or workspace. This is the supported path for keeping semantic definitions under source control alongside your transformation code.
Access
Project Roles
Workspace members are granted per-project roles: editors can change models and dashboards, viewers can only read and explore. A member with no role on a project cannot see it at all. Role assignment is covered in Team & Members.
Archive or Delete
Archiving hides a project from the workspace list while keeping its definitions intact and restorable. Deleting removes the project, its models, and its dashboards permanently, and requires confirmation by name.
