During our recent webinar, “Expand Your Automation Footprint with Intelligent Document Processing,” WorkFusion CTO Peter Cousins and Everest Group Partner Anil Vijayan discuss all things IDP: how it differs from RPA, how to integrate it into your automation program, and how it boosts automation’s impact to unprecedented levels.
Over the course of that conversation, we received more excellent questions from viewers than we had time to address. So as a follow-up, Cousins answers five of these questions in the 16-minute video above. A few excerpts:
1. Can we not train a bot in RPA to do IDP functions?
Our approach — where we have machine learning embedded at the core, and then RPA and IDP are different channels working across that same machine learning core — provides a better solution.
2. How does IDP work with fully unstructured documents like legal contracts?
The key is high-performing natural language processing.
3. How does WorkFusion handle handwritten documents?
There’s no one right answer for which engine to use. One of the things we do for customers is we select the right tool for the right job — in an automated way.
4. What is the infrastructure required to run this?
That varies. WorkFusion can run in a small environment with a handful of servers for small loads, but it’s more important that it can scale up to process extremely large volumes.
5. How about extracting data from dynamic documents like bank statements?
There’s still a structural form to them, which is leveraged by machine learning models to identify areas of the document that likely have the elements we’re looking for. What’s equally important is for the system to recognize when it sees something it doesn’t fully understand. That’s where the tight integration of human-in-the-loop…is essential — a “second pair of eyes.”