KAPTO Classification model aims to label data, fine-tune the KAPTO model behaviour on previous learning experiences to provide the best results. KAPTO model optimisation entails refining the training process using the testing phase results.    

The KAPTO Classification model serves two purposes: sort-out capability and classification.  

From the practical perspective, sort-out capabilities mean that we have content that documents produced from emails or scanned information to be converted into .PDF format. In other words, the Categorisation model aims to split the input into homogeneous pieces from the category.  

The first task of the KAPTO classification model is to understand document types. Through the Training phase, the model can read and recognise the document type to which it belongs out of these documents. This stage is categorising, understanding the belonging category and splitting the document.

KAPTO automatically and autonomously recognises the beginning and the end of the document. KAPTO model splits it into different components according to the document category defined.

KAPTO recognises and splits the received files into multiple records by assigning each part of the document to the belonging category. It is worth noting that KAPTO discards errored images. When you upload a document in KAPTO, you associate its document type with the specific case.

The AI model is trained to categorise all documents coming into KAPTO based on their category. It aims to divide the attachments into their classes and then split the document.

‍