The Evolution of Document Processing: Where We’ve Been, Where We Are Today and Where We Are Headed Tomorrow

April 25, 2022
Image of a robot's arm with emphasis on the mechanical hand.

The more you have; the more you want. Although this may not be the best motto to live by, it has many advantages when it comes to technology development. After all, some of the most innovative solutions available today are based on previous developments that were made in the past.

Intelligent document processing (IDP) grew out of optical character recognition (OCR) technology and is now serving as the foundation for the next-generation of AI-driven process automation. Let’s explore the evolution of this technology: where we’ve been, how things stand today, and the possibilities for tomorrow.

How It Began

Back in 1974, Ray Kurzweil’s company created an omni-font OCR product, which was originally created as a machine learning device for the blind. Xerox purchased the company in 1980 and began the commercialization of converting paper documents to computer text. By the 1990s, OCR had grown in popularity, digitizing historic newspapers and making inroads in the business world where manual rekeying had been the only previous method of digitizing paper-based documents.

The Development of IDP

Once OCR was established as a successful solution to digitize documents, more advanced developments would shortly follow. IDP uses OCR as a foundation to further refine both the types of documents that could be recognized and processed as well as what could be done with the extracted data.

IDP takes the best of OCR technology and applies AI and machine learning technologies to expand the number and types of documents it can process. Instead of working with only highly structured documents that OCR processed, IDP technologies are venturing out to handle semi-structured, unstructured and even hand-written text. It is also able to not only capture the content but extract the context for further data processing. IDP solutions are now evolving to handle a multitude of documents for a wide variety of companies across all different industries.

The Difference Between OCR and IDP

The main difference between OCR and IDP lies in the results. In many modern applications, OCR can only successfully process 10 percent to 15 percent of documents successfully, routing the majority of work to humans who need to manually extract, validate and structure the data. IDP, on the other hand, can fully automate the processing of 80 percent to 85 percent of the documents, flagging only rare anomalies to humans to handle.

IDP successfully extends the power of process automation, which is particularly important in high document volume industries such as financial services and insurance. For example, successfully extracting claims data allows insurance companies to handle growing volumes effectively, efficiently, and affordably.

What’s on the Horizon

Just as IDP built on the foundation of OCR, addressing its limitations and challenges, next-generation technologies will do the same for IDP. 

At Roots Automation, we’ve taken the next step with our Digital Coworker, who can not only recognize text like OCR solutions and extract data like IDP but can be trained on a specific domain to read, think and do like a human being. For example, a Digital Coworker trained for the insurance industry can recognize a myriad of documents including identification cards, claims forms, and policy documents; make accurate decisions about the data itself; be an integrated partner alongside a management suite; and drive process automation throughout an organization. 

Here’s a case-in-point. A large, growing insurance carrier faced federal regulations requiring acknowledgement of new claims within a certain time period. As volumes overwhelmed this carrier’s small and understaffed mailroom, the company engaged Roots Automation to build a Digital Coworker to help. The solution mapped incoming claims and forwarded state-appropriate acknowledgements for 99 percent of the cases with no human interaction. As a result, the carrier increased throughput volume by 60 percent and achieved a 246 percent return on investment within six months.

You can read the full FNoL automation case study here, or download multiple automation case studies here.

Many financial services, health care, and insurance organizations have already brought our Digital Coworkers onto their teams, experiencing next-level document processing today.

Read more about how our Document Vision solution can improve your document processing cycle time, reduce cost and free your people from mundane work.

Demo a Digital Coworker On-Demand

Artificial Intelligence and automation can be difficult to conceptualize.

Access our on-demand demo and watch as a Digital Coworker learns to underwrite an insurance policy, eventually reaching a 95% straight-through processing rate.
Take me to the on-demand demo
How Medical Claims Processing Transforms Patient Healthcare
Read More >
Balancing Employee Engagement and Digital Disruption with RPA
Read More >
Letter from the Co-founders: Announcing Iris, a New SaaS from Roots Automation
Read More >

Subscribe to our blog