Back to ABBYY Blog

Driving intelligent document processing opportunities with RPA vendors

Anthony Macciola

January 26, 2018

Driving content intelligence opportunities with RPA vendors | ABBYY Blog Post

Robotic process automation is one of the more promising and exciting tools in an organization’s digital transformation tool bag. Early uses have delivered a strategic impact on meaningful business processes innovation, and as a result, it’s getting C-suite level interest.

As a result, RPA vendors are growing incredibly fast with many being venture capital funded. In 2016, there was a record $5 billion VC funds flowing into AI companies worldwide, an increase of 60 percent compared to 2015. This has impacted the way information management companies collaborate with RPA providers. But first, a few challenges must be overcome before the opportunities are realized.

Not Invented Here Syndrome

While most RPA vendors are focused on their own IP, they still need additional capabilities to enable robust information, document and data capture and seamless access and integration into enterprises’ applications. However, many cannot increase their resources fast enough to deal with growing demand, and consequently, are unable to allocate resources to learn about complimentary technologies.

Compounding this challenge is the common “not invented here” syndrome where, for a variety of reasons, they prefer technologies organically grown rather than best-of-breed, proven technologies.

When partnering with an RPA vendor, it’s important to fully comprehend the various levels of RPA capabilities to ensure the best resources are utilized.

Currently, RPA solutions are moving rapidly past the screen scraping roots of the industry into highly complex value-added solutions that drive real business benefits. Document classification and automated data extraction of structured and semi-structured documents (accounts payable automation for example) are seen as high value RPA use cases.

Also, more advanced organizations have begun to look for entity extraction capabilities so they can build robots that can process unstructured content. This helps companies be more responsive and automated related to customer inquiries and process decisions. Many vendors and customers describe such advanced capabilities as “cognitive.”

There is also a lot at stake with investing in RPA. According to the 2017 McKinsey Global Institute report, “A Future that Works: Automation, Employment and Productivity,” labor associated with technically automatable activities consists of 1.1 billion FTEs worldwide, and the wages associated with technically automatable activities is $15.8 trillion worldwide. With this in mind, think of RPA use cases as three steps on a stairway.

Basic RPA

The first step is the orchestration of basic repetitive tasks. In many instances, corporate business users are creating robots without any prior training and having a positive impact on productivity and efficiency. Customers require no vendor support, professional services or even IT. RPA has engendered a self-service mindset relative to orchestration and automation.

Basic RPA includes macro-based applets, screen scraping data collection, workflow automation, Visio-type building blocks, process mapping and business process management. They eliminate the swivel chair processes for data entry commonly seen in logistics and transportation and invoice processing.

Enhanced RPA

Use cases in the second step include the extraction of metadata from forms, followed the inclusion of completely unstructured content within RPA use cases such as contracts. Enhanced RPA addresses automation of processes that are less structured and often more specialized. Tools and platforms supporting enhanced process automation offer some capabilities such as out-of-the-box built-in knowledge, an understanding of natural language, ability to consume and leverage unstructured data, automated learning capability, pattern recognition and e-bonding capabilities to other well-established software platforms.

Enhanced process automation does require a stronger feedback mechanism to leverage the learning capabilities built around machine learning to continually increase savings, so having the right skills on hand is essential to continuous improvement.

Cognitive RPA

The third step in the RPA stairway includes learning of human interaction with RPA robots and the eventual recommendation and automation of action, that enable gaining better understanding of the impact of automation within an organization.

Cognitive RPA combines advanced technologies such as natural language processing, artificial intelligence, machine learning and data analytics to mimic human activities such as perceiving, inferring, gathering evidence, hypothesizing, reasoning and interacting with human counterparts. Envision the capabilities in self-driving vehicles where systems are taught rather than programmed, a process that can take months to years depending on the complexity of the problem domain.

Cognitive RPA requires the largest investment in time and dollars but has the greatest potential to transform. Therefore, it will require new hires or “rented” expertise as this expertise is rarely found internally. Industries such as oil and gas or financial services benefit most from cognitive automation because of a high quantity of available data, tight regulation and continuous improvement.

Opportunities with RPA Vendors

The smart RPA vendors will realize they can’t be everything to everyone. Their respective roadmaps are considerable so the likelihood of them wanting to recreate complimentary technologies is minimal. The vendors that have higher ‘not invented here’ mentalities will likely experience a higher failure rate when trying to integrate on their own.

Demand for the more advanced cognitive uses cases will grow dramatically in 2018 and will start with structured and semi-structured documents but will quickly move to include completely unstructured documents. It is essential to be prepared to illustrate your expertise to compliment and meet RPA demand on multiple levels.

This is a slightly abridged version of Anthony Macciola's article that appeared in Information To read the full-length version, please click here:

Intelligent Document Processing (IDP) Digital Transformation Artificial Intelligence (AI) Robotic Process Automation (RPA)
Anthony Macciola ABBYY

Anthony Macciola

Subscribe for blog updates


Connect with us