Back to Case Studies
Online Textile StoreTextile

Order verification system with barcode scanning — saving ₪19,000 per month

The Challenge

An Israeli textile store selling online and managing an independent warehouse with hundreds of items in various variations — sizes, colors, fabric types, and models. Approximately 200 orders per day, about 6,000 per month, a team of 3 warehouse workers, and a picking process entirely based on printed sheets.

The issue wasn't the volume of work but its accuracy. When a warehouse worker searches for a product by name on a printed sheet, without any verification mechanism — mistakes are inevitable.

Main Issues

  • Manual picking without verification — the worker reads a product name from a printed sheet, searches on the shelf, and places it in a bag. There is no step to check that the product taken is indeed the product ordered
  • Confusion between variations — a single-size sheet in cream looks almost identical to a double sheet in white. A towel 50x100 and 70x140 in the same color? Impossible to distinguish by eye
  • Late discovery — the mistake is only discovered when the customer opens the package at home. By then, shipping fees have been paid, and the business needs to pay again for return shipping and reshipping the correct product
  • High cost for each mistake — return shipping, reshipping, employee time to handle the complaint, repackaging, and sometimes a product that returns in a condition that cannot be resold. Average cost per mistake: ₪150-200
  • Inaccurate inventory — without digital tracking of what was taken from the shelf, inventory counts always showed discrepancies. Products that the system showed as in stock were not on the shelf, and vice versa

₪19,000

Monthly cost of packaging errors

~120

Incorrect orders per month

Out of 6,000 orders per month, about 2% were sent with an error — approximately 120 orders. Each one was a disappointed customer, a return to handle, and a cost absorbed into the profit.

Our Solution

Day2 AI has developed an intelligent system that combines barcode scanning with artificial intelligence to eliminate packaging errors. The system runs on the warehouse worker's phone, completely replacing printed sheets.

Order Import with AI

Orders are automatically imported into the system from Shopify — including all details: product name, barcode, size, color, and quantity.

But the real advantage is with orders that arrive as PDF files via email — such as orders from networks or suppliers. An AI engine automatically analyzes the PDF, identifies the order details (product name, quantity, size, SKU), breaks them down into separate items, and adds them to the system — without any manual typing. The AI can handle different order formats and learns to improve over time.

Each order is displayed in a clear interface with status: pending, in process, or completed.

Order Management Interface

Order Management Interface — Real-time status of each order

Smart Scanning and Verification

The worker opens an order and clicks the scan button. The phone's camera opens as a barcode scanner. The worker points the camera at the product — and the system immediately verifies:

  • Green mark if the product matches the order
  • Yellow alert if the barcode is unknown — with an option to add manually
  • Red block if the product does not belong to this order

Each scan is accompanied by a phone vibration, so the worker feels the confirmation without looking at the screen. Items without a barcode — such as small accessories — can be marked manually.

Barcode Scanning Screen

Scanning Screen — Immediate confirmation after barcode scan

Final Verification with AI

When all items have been scanned, the worker clicks the "Finish — Verify Order" button. Here, an additional AI stage is introduced — the system performs a smart verification that checks not only that every item has been scanned, but also:

  • Whether the quantities exactly match the original order
  • Whether there are any excess items mistakenly included
  • Whether there is a mismatch between the barcode and the ordered variation

If everything is correct — the order is marked as "completed" and ready for shipment. If there is a problem — the system displays a detailed list of what is missing or incorrect, and the worker can return and correct it before the package leaves the warehouse.

Before and After

Before After
Orders per day ~200 ~200
Picking method Printed sheet and eyes Barcode scanning + AI
PDF order import Manual typing Automatic decoding with AI
Product verification None Automatic, real-time
Errors per month ~120 (2%) ~6 (0.1%)
Monthly error cost ₪19,000 ~₪900
When error is discovered At the customer Before the package leaves

Net Monthly Savings

₪19,000

The investment in the system paid off within two weeks

Measurable Results

Quantifiable outcomes that demonstrate the impact of our solution

Monthly savings
₪19,000

Complete savings on packaging errors — returns, duplicate shipments, and team time

Reduction in errors
95%

From 120 incorrect orders per month to just 6

Faster picking
40%

Digital scanning + AI is faster than manual search

Return on Investment
Two weeks

The system paid for itself within two weeks from the day of activation.

Before the system, we handled dozens of returns a week. Today, it's a number you can count on one hand. The employees adapted within a day — they scan, the phone confirms, and the package is sent. What surprised us the most was how quickly it changed the experience for our customers.
Operations Manager, Online Textile Store

Ready to Achieve Similar Results?

Let's discuss how Day2 AI can transform your textile operations with AI automation.