Content Intelligence to Fuel Cognitive Automation

Good organizations become great by investing in digital transformation, adopting technologies and solutions that focus on enabling them to adapt to the demands of customers and the markets they serve. Digital transformation requires completing the digitization of internal and external business processes through which customer experience and operational efficiency can be significantly enhanced. At the center of any digital transformation strategy is the organization’s content, whether it’s paper being converted into digital formats and connected to processes or whether it’s digitally born customer communications, correspondence, forms, social media or other digital documents.

The term digital transformation is broad in nature and encompasses many automation technologies. I’d like to explore one, robotic process automation (RPA), that has exploded onto the market. RPA is helping organizations to eliminate manual processes by using intelligent software robots that can automate a wide array of manual human tasks and activities by simply mimicking what a user does across one or many applications. Quite often these tasks and activities involve the processing of both structured and unstructured content.

To continue to fuel the use and expansion of RPA and the value it delivers, organizations using RPA will need to look toward technologies and solutions that deliver an added level of content intelligence, to make the robots smarter and more effective.

Intelligent Robot Fundamentals

A software robot needs to be smart in a number of ways: First, it must be able to follow a set path through one or more applications for inputting, outputting and validating data across many systems. Secondly, it must be able to understand the data and documents it’s processing, like invoices, contracts, financial documents, customer memos and more. To achieve the second goal, RPA must rely on best-of-breed intelligent content services that provide the robot with machine learning, natural language processing and contextual understanding of content capabilities that enable robots to automatically classify and extract data from content, reducing or in some cases eliminating the need for human involvement. A wide array of use cases for RPA span different classes of intelligent process automation,

Intelligent Automation Spectrum

There are varying degrees of automation where RPA and cognitive platforms come together to address three classes of intelligent process automation: basic automation, enhanced robotic process automation and cognitive automation. Basic robotic process automation can make sense in scenarios where the process activities are simple rules-driven work, are repeatable and represent medium to high transaction volume and where the business case will deliver extremely good ROI by eliminating humans from doing remedial tasks. Any ineffective process involving humans , entering and copying data, and reconciling results is a perfect place for organizations to start. In basic automation, there’s often content (text documents, images, PDFs, etc.) involved as well, but the level of complexity in terms of content variety and required task is relatively low. Processing content in basic automation involves scraping data from a screen (e.g., application within a Citrix environment) or handling a simple task like performing OCR on a document and converting to PDF to make it searchable and actionable.

The next class is enhanced robotic process automation, which builds on the same drivers as basic automation but has a stronger focus on processes involving structured and unstructured data. In this scenario, a robot to drive down the amount of time humans spend on the process.

might look like the following: A robot receives an email and uses content intelligence services to understand the specifics of the request and attached documents; it connects to one or many systems to process the request and notifies the customer when the request has been completed. In this example, the process cannot happen without the “intelligence” of content services to classify and extract the data; otherwise, humans must manually read through documents and extract data. Across many industries, the development of a robot encompasses not only automating the actions a user takes with an application, but also classifying the content and using that information to drive the process forward.

The Road Toward Cognitive Automation

The third and final class is focused on using advanced technologies – natural language processing, predictive analytics and more that further advance the use of content intelligence services – that fall into the category of artificial intelligence or cognitive automation. In this class, robots can apply text analysis to identify entities; understand facts and events accurately; build stories across documents; detect the relationship between entities, such as who is the seller or the buyer in a contract; and classify large volumes of unstructured data.

Building Robots for Today and the Future

The opportunity surrounding RPA is fast-moving, with hundreds of possible use cases in an enterprise organization, many of which require processing content. The key to success is first to focus on the problem(s) you want to solve and ensure you have the right technology to handle those content-centric RPA use cases. Then, consider the longer-term strategy around RPA and build your RPA practice with a vision toward . You don’t need to start off building the smartest robots, but having the right intelligent robot technology foundation in place will certainly prepare you for the future.