I apologize in advance if I asked an incorrect question here. But do not judge strictly, just tell me where to go or give a link to information.

The essence of my question is the following: I have several types of documents (receipt note, invoice, payment receipt, etc.). I pass these documents through OCR to parse and enter the necessary data into the database.

However, situations are increasingly common when invoices come from different suppliers (these can be suppliers already known in the database, or they can be new suppliers) and the invoices template is different. That is, the document type is the same – it is an invoice, but the templates of this type of document is different for different companies. And this creates difficulties with entering information into the database. That is, I need to create some algorithm that will determine the template of the document (template from which company? new template or existing? and so on)

Naturally, I am not asking for code, I am asking for links where I can read the relevant information to solve this problem or a resource where I can ask more specific questions on this topic