NETWORK - SmartPDF - Optimized data capture processing flow

SmartPDF combines three data capture technologies: Optical Character Recognition (OCR), machine-readable PDF layout mapping, and AI. Before, these were offered as two separate services (Scan and Capture and PDF e-invoice). SmartPDF provides easy-to-use optimized data capture processes.

The business documents are first routed to the Optical Character Recognition (OCR) or AI flow, and if the document is a machine-readable PDF, the mapping countdown will start unless AI processes the document. The most optimized data capture process is where AI automatically processes the business document (invoice/credit invoice). AI capabilities are developed continuously.

Requirements for mapping:
- The PDF must be machine-readable
- 5 business documents (layout) processed in a single calendar month

A business document can be an invoice or credit invoice, and several different suppliers can send it to count towards a minimum of 5 processed documents per layout. The Basware's system will automatically flag such layout, which is eligible for layout mapping. All customers using SmartPDF are sharing the layout pool. Once the layout is created as a template, the receiver side custom mask needs to be adjusted, especially if feedback is given for previously processed documents.

The layout is typically related to the invoicing system (ERP). Layouts might have company-specific changes even if two suppliers use the same invoicing system. The mapping is done on the metadata side, which is not visible to the human eyes, meaning there is structured data behind the PDF file. The two documents might seem to have the same layout with human eyes, but the visible version is a visualization of the structured data that could be differently constructed.

Customer's using Network P2P:

Here is in detail how P2P can show the origin service:

In P2P: Accounts payable -> Invoices -> Advanced search -> Add criteria -> Origin Service

For SmartPDF origins
SmartPDF - AI = Automatic capturing (Gateway) AI technology
- AI will process all PDF types as the AI has built-in an OCR machine to read image-based PDF files (non-machine-readable PDF files). It cannot be activated if custom fields and/or line levels.
SmartPDF - template = Automatic capturing (Gateway) technology
- Same document layouts are shared with all SmartPDF customers where processing is via templates, to identify a layout, there needs to be at least once 5 documents/month to trigger template creation. The supplier template is mapped for a document layout, and buyer-side requirements are created as a receiver template, which filters captured data per the receiver's customized needs.
SmartPDF - OCR = Manual data processing via human validation partner
- Where OCR reads different types of PDF files (machine-readable PDF files and image-based PDF files (non-machine-readable PDF files), populating suggestions that a human validation specialist reviews and makes adjustments. Data processing is +24 hrs. and similar to the Scan and Capture process. Automatic processing will eat the OCR volume if the criteria are met, and the Gateway team is constantly analyzing and creating new layouts.
SmartPDF - self-validation = Manual data processing via the buyer’s own team
- Requires a separate SmartPDF setup similar to the CloudScan service, however, the automatic processing is done with the AI.

Other origin services:
Scan and Capture
Scan and Capture – email
Scan and Capture – paper
Scan and Capture - self-scanned
Scanned by Basware CloudScan
CloudScan with Basware-validation