Our AI model learns from past orders to automatically fill in data from scanned order confirmation documents into the order form. The AI model fills in suggestions for both New orders, but also orders that are created via a pre-existing order template. In both instances, the system will scan a pdf order confirmation document and prefill any suggestions.
Here’s a detailed overview of how it works and what you can expect:
How it works
1. Learning Process
The AI model learns every time a user creates an order for a specific customer using a scanned document.
Typically, it takes about 3-5 scanned orders for the model to start providing pre-fills.
2. Automatic Pre-Fills
When the model is confident that a value in the document matches a field in the order form, it will automatically fill in the corresponding field.
Users will receive a notification in the top-right corner of the screen when one or more fields are pre-filled.
3. User Interface
The recognised fields include:
Customer by VAT Number - must be added to the customer profile
Customer name - option to create alias with matched names
References
Dates / Times
Opening times
Numbers
Addresses
Products / Packaging
Descriptions / Additional instructions
Collection / Delivery Notes
Good descriptions
Click on a field in the order and a dropdown will open.
Some options in that dropdown will have a label 'Suggestion'.
This is data that is fetched from the uploaded document.
Pre-filled fields will be highlighted in blue within the form, similar to when users manually select a suggestion from the document.
Verify accuracy of the pre-filled data and manually complete any fields that are not automatically filled.
📚 If NOTES do not pre-fill typing in the box will display any matching options from the document.
4. Accuracy and Limitations
The model is designed to provide pre-fills only when it is nearly 100% certain of the accuracy, aiming to minimise incorrect pre-fills.
Users should not expect the entire form to be pre-filled but can anticipate a significant reduction in manual data entry.
📚 Document scanner will search for numeric values, addresses, dates/times and products to use as suggestions.
Using a template order
It is possible to use the document scanning and prefill functions alongside an order template.
Navigate to ORDERS > Templates
Select the template to copy and click 'copy selected orders'
From the new order that is created, click 'Edit order'
This opens the edit order page, including a section where the transport service and the upload documentation fields are available.
From the 'Order confirmation from customer' section, use the 'drag & drop' feature, or click to upload a file.
The system will begin scanning and adding matched suggestions, shown highlighted in blue.
Verify the copied information and enter any missing details.
Click SAVE.
Click Play👇 to watch an overview of how to use the customer order confirmation document to create a new order from a template.
Split large PDF files for Scanning
Create an order by dragging a pdf into Qargo. It’s possible to define which pages of the PDF file(s) should be used for the order pre-fill and as order confirmation.
User can specify what page range should be used. This range will be used to scan by AI and it will be used as the order confirmation attachment in the order.
Once the split has been created, this will be visible to review as well.
Create splits for the different PDF files that have been attached.
The original PDF files can be deleted or they can be left to attach to the order as well.
📚 When multiple splits are created, the last created split will be used as order confirmation.
The last created split will be added in attachment as the order confirmation from the customer as well as the email itself.
Other splits created from other PDF files will be added in attachment as document type ‘Other’.
The original PDF files are added at document type ‘Communication’
Customers can now easily extract the order confirmation from larger PDF files.
More accurate document scanning by specifying pages to be used.