Importing data overview

Importing data is a multi-step process that maps to the first two stages of the Medallion architecture:

Bronze (raw)         →    Silver
Import files into         Create your dataset
a batch                   from the batch

Stage 1 is fast — you upload CSV / Excel files and run them through. Stage 2 is where the structure work happens — you build a dataset module that becomes part of your project.

When to import

Typical times you’ll start an import:

• A new project has just been created and has no data yet.
• A customer has sent you updated source files.
• An earlier import had issues and you need to redo it.

The flow at a glance

Stage 1 — Import files (raw / bronze)

1. Create an import batch — sits inside the project, groups the files you’re about to upload.
2. Upload CSV / Excel files into the batch.
3. Run the import — RebelCore™ parses the files. The batch moves to status Created.

Stage 2 — Create your dataset (silver)

4. Open the Module Dataset Builder for the batch.
5. Pick the hierarchy / labels, review the AI Data Advisor’s suggestions.
6. Click BUILD MODULE — the dataset is now part of your project.

After stage 2 the dataset shows up in the Tree, where you do the gold-tier curation. Each step is incremental — you can leave a batch half-done, come back tomorrow, and pick it up.

Caution

Importing files alone isn’t enough — until you Create your dataset from the batch, the data won’t appear in the Tree. The CREATE DATASET action on the Datasets page is what links an upload to a project.

Tracking status

Open the Imports page from the side menu (it shows the heading Datasets) to see all batches and their status:

• Draft — batch exists, files may or may not be uploaded yet.
• Processing — RebelCore™ is parsing the uploaded files.
• Created — files parsed; batch is ready to be built into a dataset (the CREATE DATASET action appears).
• Built / linked — dataset is part of a project. The Dataset Availability column shows its label path.
• Failed — something went wrong; click the batch for details.

The page keeps the table visible while it refreshes, so you can continue reading the current list instead of watching it blank out. Use Refresh when you want to manually check for the latest import or dataset-build status.

What’s next

When you’re ready to do an import end-to-end:

1. Import files — bronze stage, get the data in.
2. Create your dataset — silver stage, give it structure.
3. Explore the Tree — gold stage, select what matters.