Skip to main content

How AIDA works

AIDA has been conceived and designed so that its use can be simple and immediate for all users, giving them the possibility to build their workflow step by step, allowing them to modify and adapt the structure created according to their needs at any time.

Creation of the data model (data-driven)

The first basic action to be taken is to create a new document type. AIDA will guide you to the creation of the document type starting with property groups.

Before you begin, however, you should bear three elements in mind:

  • What type of document do you want to analyse: for example, a purchase invoice, a transport document, a bill, a contract, a receipt, an event ticket or a transport ticket.
    Remember that the type is not linked to the graphic aspect (layout/template) of the document, but only to the properties you wish to extract from it.

  • What information (properties) to extract from it: taking the invoice as an example, you may need to extract the date of issue, the company issuing it, the VAT number, the total amount...

  • What to do with the document once it has been analysed: for example, index the document in the AIDA archive or send it outside AIDA via a Destination.

Once the desired document type has been created, you can move on to configure all the properties to be extracted from the document.

Intelligent document analysis

The analysis section allows you to train AIDA with a few clicks and in an adaptive way to automatically extract properties from your documents.

Each document processed by AIDA undergoes an OCR phase, which is the first step in its intelligent analysis. AIDA's OCR engine also supports handwritten text and documents rotated around the vertical or horizontal axis.

Click on a pending document to enter the training section: here, simply select the desired property in the left-hand column, and then click on the property value within the document on the right. AIDA will take care of all conversions of values ( for example, to dates or numbers) or corrections, such as removal of extraneous text, for you.

Training is only finalised when you Confirm the document by clicking on the button in the top right of the screen: until then, training will only be in draft form.

Confirmation also allows you to associate activities with the document, send it to archive or to a Destination, and finally to create an automatism that can perform the approval action automatically.

Remember that if you have documents awaiting approval, and you train AIDA on other documents in the meantime, the former will not benefit from the new training information: this is because they were processed when this information had not yet been entered into the AIDA engine.

However, it is sufficient to reanalyse them so that they can benefit from the latest learning information.

AIDA is able to detect and report anomalies that may occur when analysing the document, such anomalies will be highlighted in yellow: in these cases it is sufficient to check the value and confirm it, if correct, or correct it, if incorrect.

Inputs / Destinations document flows

AIDA processes new documents through Inputs; incoming documents are placed in one queue and processed as quickly as possible.

AIDA accepts and supports dozens of different incoming file types and converts them for you into PDF, the document standard.

Documents can be sent externally via Destinations. It is possible to choose one or more destinations when approving the document, thus allowing multiple outputs, according to the rules you configure.

You can also export document property values via the archive.

Supported languages and limitations

AIDA supports text capture in English, Italian, French, German, Spanish and Portuguese.

AIDA supports both printed and handwritten text.

The maximum number of pages for a single PDF document is 2500, while the maximum file size of the single document is 40 MB.

Supported file types

AIDA accepts the following file types as input and converts them to PDF:

FormatExtensionWeb uploadE-mail attachment
BitmapBMP
Comma-separated valuesCSV
Microsoft WordDOC - DOT - DOCX
AutoCAD DXFDXF
Encapsulated PostScriptEPS
Graphics Interchange FormatGIF*
HTMLHTML - HTM
JPEGJPG - JPEG*
Outlook emailMSG
Portable Document FormatPDF
Digital signature PAdES / CAdESPDF - P7M
Portable Network GraphicPNG*
Microsoft PowerPointPPT - PPS - POT - PPTX
PhotoShopPSD
Microsoft Pocket WordPSW
Microsoft Pocket ExcelPXL
Rich Text FormatRTF
SGVSGV
Scalable Vector GraphicsSVG
OpenOffice.org XMLSXW - STW - SXC - STC - SXI - STI - SXD - STD - SXM
Truevision TGA (Targa)TGA
Tagged Image File FormatTIF - TIFF
Plain textTXT
Microsoft VisioVSD
Microsoft ExcelXLS - XLW - XLT - XLSX
XMLXML
ZIPZIP
  • It is possible to disable JPG/PNG/GIF formats, or to require a minimum size, with the e-mail input configuration.