New Import Wizard

Queue a million-row import job from the wizard.

This flow is positioned for very large Excel, CSV, and JSON imports. The browser only submits the job definition while validation, queueing, and write work move into the background.

Wizard Steps

Eight steps

The spec-defined flow stays explicit because million-row imports need predictable staging, not a one-shot browser upload.

Step 01

Choose entity

Products, Customers, Orders

Step 02

Choose format

XLS, XLSX, CSV, JSON

Step 03

Upload file

File metadata stored locally

Step 04

Select worksheet or root

Sheet name or JSON root

Step 05

Header mapping

Template-backed mapping

Step 06

Validation preview

Warnings vs blocking issues

Step 07

Import options

Match, write mode, dry run

Step 08

Run in background

Creates a persisted demo job

Create Import Job

Submit an import to the demo queue

File uploads are used for metadata only in this prototype. The job, options, logs, artifacts, and timeline are persisted so the app can demonstrate a high-volume, background-first import posture.

Validation Preview

What the tester should expect

Each submitted import starts with queue or validation state and carries its options, file metadata, logs, and starter artifacts into the Jobs surface, reinforcing that large imports are meant to run off the main browser path.

Parsed rows

Derived from the estimated row count, including million-row planning

Blocking errors

Seeded for order-mode warnings where relevant

Warnings

Stored as row-level issues on the job instead of getting lost in a giant file

Artifacts

Validation, canonical, errors, warnings, logs

Workflow	Behavior
Validate only	Creates a job in validating state so the preview path can be exercised.
Validate + import	Creates a queued import ready for manual testing in the queue UI, framed as a background job for very large datasets.
Historical orders	Seeds a warning about best-effort discount parity in the job detail view.