Skip to content
RebelCore Docs

How the Tree works

The Tree is the main workspace inside a project. It presents the project’s completed modules as a navigable hierarchy so you can choose the data relevant to a task.

Open the Tree

  1. Open Projects under Workspace.
  2. Select a project card.
  3. Expand the Tree branches to explore the available data.

You need View Tree permission and access to the project.

Read the hierarchy

The Tree can include:

  • Dataset and module branches.
  • Labels assigned during module creation.
  • Worksheets or source groupings.
  • Individual fields at the leaves.

Projects with multiple modules display all selected modules in one Tree, organised by their labels. A module can appear in more than one labelled area when it was assigned to more than one business context.

Use the search field to find a node by name.

Select data

Tick or untick leaves to control which fields are included in the current selection. Parent branches help you select or review related fields together.

Aim to include:

  • Fields directly relevant to the intended Agent questions.
  • Descriptive fields needed to interpret the result.
  • Measures and dates required for the analysis.

Avoid including unrelated identifiers, notes, or duplicate fields merely because they are available.

Use the blade

Select a node to open the blade on the right.

TabPurpose
DETAILSReview the selected node, its path, type, selection weight, and variable count.
CHATReview and apply AI Data Advisor suggestions, where available.
DATATrain, open, and manage semantic datasets for the project.

The tabs shown can depend on the node and your permissions.

Review suggestions

In CHAT, RebelCore can suggest fields that may be useful for common questions. Review each suggestion before applying it. Suggestions change the current selection; they do not change the uploaded source file.

See Tree chat & AI suggestions.

Train a model

The DATA tab turns the current selection into a semantic dataset that Agent can use.

  1. Confirm the selected fields.
  2. Open DATA.
  3. Enter a clear model name.
  4. Click Train Model.
  5. Wait for the status to become READY.

See Semantic dataset.

Open Agent

When a trained model is ready, select Question in RebelCore Agent. Agent opens with the project and model selected.

You need Access Agent permission. If the action is missing, ask your administrator to review your role.

Troubleshooting

  • A module is missing: confirm that it was selected when the project was created.
  • A field is missing: review the source import and module build.
  • The Tree is empty: confirm that the selected module completed successfully.
  • Train Model is unavailable: confirm that fields are selected, a name is entered, and your role has Create Semantic Dataset.
  • Agent cannot be opened: confirm that the model is READY and you have Access Agent.