Does anyone have any experience with Intelligent Document Processing technology or any providers that offer this technology/tool?
Sort by:
Thanks Steve!
If you could please define more exactly what type of processing you are looking for, that would help to see, who has what experience in the community.
Example: are you looking for RPA type document processing or something else or more, like RPA supported full document management systems?
Do you need reading and processing content of hard copy documents to file for repository, or to also categorise and analyse etc?
The point is, what do you need from the solution (what are your requirements)?
Thank you!
Here are our requirements:<br><br>Improve document handling processes using an Intelligent Document Processing (IDP) solution that reduces human intervention by leveraging advanced OCR & AI technologies.<br><br>1. Automate document classification, i.e. auto-classify documents by category (supplier invoice, export document, packing list, ….) by use of Machine learning<br>2. Split PDF’s containing multiple type of documents into multiple PDF’s and auto-classify the documents<br>3. Reduce manual data extraction, at increased velocity, and with increased accuracy without the need to setup OCR templates<br>•Auto-identify pre-defined required data elements on each document type (Machine learning)<br>•Compute confidence level per identified data element (i.e. how sure is the system that the data point is correctly identified) (Machine learning)<br>•Create work queue for data points not meeting set confidence level threshold (i.e. manual intervention needed to validate what system has identified)<br>•Extract identified data points targeting further downstream processing within Caterpillar<br>•Leverage global team to train the “common global brain” instead of having multiple instances of the “brain” by transportation team, leading to duplication of efforts<br>4. Real time performance dashboards<br>•Perform analytics on processed data to further improve insight into metrics produced by the system during execution (e.g. analyze mistakes & define thresholds)<br>5. Digitization of document content enables for integration of document data in a DWH and store the digital version as facts (occurrences) identified against dimensions (e.g. doc types, suppliers, carriers, brokers, etc.)
We achieved some work with IDP. And we have plan to increase its usage for more processes.
Captiva (now Intelligent Capture), from OpenText, is the solution we're using. Is it perfect? No, but it works quite well.
We're looking to do some experiment with RPA that also have IDP capabilities because sometimes, capturing a document also means generating actions in some applications.
And sometimes, once the document is captured, we only store it with some metadata.