FlexiCapture 12 deploys Machine Learning

ABBYY’s FlexiCapture 12 has been enhanced with machine learning (ML) technology to improve the accuracy and speed of processing enterprise content.

“Powered with ML and AI, ABBYY FlexiCapture is the most comprehensive, robust, and technologically advanced platform that supports organizations’ digital strategies,” commented Bruce Orcutt, Senior Vice President of Product Marketing at ABBYY.

Additionally, ABBYY is testing the application of natural language processing (NLP) technology in FlexiCapture to extend capture capabilities to unstructured documents. NLP will make unstructured data easier to understand, analyse and consume by providing structure through content classification and entity extraction.

Through the use of ML and AI, internal ABBYY tests demonstrated that FlexiCapture 12.3 can achieve a more than 93% extraction accuracy rate that can feed data into robotic process automation (RPA), business process management (BPM) and content driven process engines.

Other performance enhancements include:

  • Overall FlexiCapture 12.3 platform performance is improved by 50% for small batches (3-5 pages) and 10% for large batches (100 pages)
  • Updated neural network improved invoice extraction quality
  • New machine learning methods increased extraction rates and improved processing of complex tables.