DOCUN is a document solution that finds matching information for the item to be extracted for non-standardized documents.
As our solution supports various document formats (pdf, hwp, doc, etc.) and image files by optimizing the models(Text Extracting/detection and recognition model, Key-value Extraction model), the personnel who inspect/correct only 10% of the total key-value items are needed. As a result, our solution reduces the required input manpower by 90% and shows higher accuracy over time through additional learning on error/exception items.
Jenti Inc.
PROJECT
DOCUN(DOCument UNderstanding) DOCument UNderstanding solution for extracting key-value information at unstructured information
