Skip to main content
The ABBYY FlexiCapture classification feature can be used to implement the following scenarios:
  • Automatic sorting of incoming documents
Example. The following two types of documents are received — invoices and contracts. ABBYY FlexiCapture automatically separates documents of one type from the other. The sorted documents are then forwarded to the appropriate departments responsible for their processing. Invoices should not be forwarded to the legal department, as this will distract the lawyers from contracts and delay the processing of the misplaced invoices by the accounting department.
  • Assembly of page images into documents followed by automatic sorting
Example. Scanned document pages are fed to ABBYY FlexiCapture. The pages are not assembled into documents during scanning. Instead, the classification feature is used to classify the pages. Then ABBYY FlexiCapture assembles the classified pages into documents and determines the type of each document.
Note that no classification technology is 100% accurate, so documents classified with low-confidence have to be checked by a human operator. As an alternative solution, you can capture data from low-confidence documents and perform data checks. If the checks are passed, the documents will be deemed to have been classified correctly.
  • Detection of documents that need to be processed
Example. A PDF file is fed to ABBYY FlexiCapture which contains various documents related to a particular case, such as a mortgage application, and ABBYY FlexiCapture detects documents of certain types (e.g. passport, application form, bank statement, etc.). Classification is used to quickly find the relevant pages from which data should be extracted. The remaining pages are saved to a searchable PDF file.
  • Processing support requests
Example. A bank receives various types of support requests from its customers. Support requests are automatically classified based on the their content (new payment details, bounced payment, loss of a credit card, etc.). Urgent requests are separated from non-urgent ones and forwarded to their respective managers. Additionally, those requests are identified which an automatic reply can be sent.