Walter AI Feature

Walter Rosetta: Bulk Import

Import orders from any CSV, Excel, PDF, or JSON file. Walter AI automatically maps your columns to the WarpWare order schema — no manual configuration required.

What Rosetta Does

You upload
Any order export file — CSV, Excel, PDF, or JSON from any platform
Walter analyzes
AI reads your column headers and sample data, maps them to WarpWare fields
Orders created
After you review and confirm, orders flow directly into the fulfillment pipeline
Supported file formats
.csv
Standard CSV
.xlsx / .xls
Excel Spreadsheet
.pdf
Structured PDF tables
.json
JSON array or object
Import Steps

Import Orders with Rosetta

1

Prepare your file

Export your orders from your source platform — Shopify, Amazon, a spreadsheet, wherever they live. Rosetta handles messy report-style files too, but clean tabular exports import most reliably.

Required columns (at minimum):
Order ID / Number
Unique identifier per order
SKU / Item Code
Product identifier per line item
Quantity
Quantity ordered per line item
Shipping address, customer name, email, and order notes are optional but will be imported if found. The more complete your file, the better.
2

Open Walter Rosetta in WarpWare

In WarpWare, navigate to:

AssistantRosetta

Rosetta is part of the Walter AI suite. Select the sub-client you want to import orders into using the organization selector at the top of the page before uploading.

3

Upload your file — Walter analyzes it

Drag and drop or click to upload your file (up to 50 MB, 10,000 rows). Walter reads your column headers and sample data, then automatically suggests mappings to the WarpWare order schema.

What Walter shows you
Detected format: e.g. "Shopify Order Export" — confidence shown as a percentage
Column mappings: Each source column → WarpWare field, with a confidence score
Warnings: Missing required fields or ambiguous columns flagged before you import
Walter learns from every import. If your file format has been seen before, mappings are suggested with higher confidence automatically.
4

Review and adjust mappings

Review Walter's suggested column mappings. High-confidence mappings (90%+) are usually correct. Use the dropdowns to correct any that look wrong before proceeding.

High-confidence rows are safe to accept as-is
Lower-confidence rows are worth double-checking
Unmapped columns are listed separately — map them or leave them
Required fields (order ID, SKU, quantity) must be mapped to proceed
5

Preview the first 10 orders

Before committing the import, Rosetta shows you the first 10 orders as WarpWare will interpret them. Verify that names, addresses, SKUs, and quantities look correct.

If the preview looks wrong (e.g. SKU and quantity are swapped), go back and correct the mappings. The import cannot be undone once submitted.
6

Import — orders are live

Click Import to commit. Rosetta processes all rows and shows a results summary:

Created
Skipped (duplicates)
Failed

Created orders immediately enter the WarpWare pipeline, run through automation rules, and are queued for submission to your WMS.

Frequently Asked Questions

What if my file has one row per line item (not one row per order)?
Walter detects this automatically and groups rows by order ID to build multi-line-item orders correctly. You'll see the detected row structure in the analysis results before importing.
Is there a file size limit?
50 MB per file, up to 10,000 rows. For larger datasets, split the file into multiple imports.
Can I re-import the same file?
Duplicate orders (same order ID for the same client) are automatically skipped. Re-importing a file is safe — only net-new orders will be created.
What happens to orders after import?
Imported orders enter the standard WarpWare pipeline: they run through automation rules (hold logic, routing, processing rules) and are queued for submission to your connected WMS.
Does Walter learn from my corrections?
Yes. After each import, the column fingerprint and final mapping are saved to a knowledge base. Future uploads with the same column structure get better auto-suggestions.

Need help with an import?

Contact Support