Modern customer service—More social, more intelligent

Coordinating multiple RPA processes

Once a number of RPAs have been established, managing them can become a challenge. “When you have a lot of  processes, it can become like a spider’s web with tracking and maintenance issues,” said Stefan Andreasen, co-founder of SmartRPA. Governance and change management are needed in order to keep them up-to-date and eliminate conflicts. SmartRPA offers a software product called Catalyst, which is a tool to manage RPA processes. “Organizations must understand how their RPA processes are running and what the interdependencies are,” he explained. “For example, there might be five different processes that access a company’s email service. If the email server’s policies change, Catalyst can update all five RPA processes using one component.” Catalyst is not an RPA; it runs on top of RPA solutions such as UiPath, Blue Prism, and Kofax RPA, which Andreasen co-founded as Kapow and was bought by Kofax.

Selecting the right processes to automate using RPA is important, Andreasen noted. “If a process is being carried out millions of times a day, it should be coded into a primary application, and, if it is a very rare event, it can be done manually,” he said. “On the other hand, if it is relatively frequent but not high volume and uses multiple enterprise systems, RPA could be the right answer.” An example of such a scenario is when a bank customer reaches the age of 18—an event that applies to only a small percent of customers—and certain processes need to take place that can be triggered by RPA.

“One of the hurdles that is popping up along with no-code and low-code RPA solutions is that IT has not been involved enough,” observed Andreasen. “Business users are developing systems on top of enterprise systems, so IT needs to be involved. Email servers are a good example. If only business users are involved, implementation could go the wrong way. Both are needed.” Since RPA is often used in digital transformations, a broad perspective is essential. “RPA is a valuable tool to have if it is used right, and there is no doubt that its use is accelerating because of digital transformation,” Andreasen concluded.

Looking to the future

Without question, AI will increasingly be used in all aspects of customer service and in its enabling technologies. “AI will provide the thinking, and RPA provide the ‘doing,’” said Gordon Robinson, senior director of R&D in decisioning at SAS, a data analytics company that builds machine learning and other AI technologies into many of its solutions. “If you look at the number of organizations that need to drive efficiencies by automating as much as possible, you can see the critical role of RPA in digital transformation.”

The increased intelligence will come both from making RPA applications more intelligent, and from being able to ingest and act on information from cognitive systems such as customer sentiment analysis. “We are embedding machine learning and other AI capabilities into RPA [processes] to make them smarter,” noted Robinson. “The two are very complementary.”

These trends can be expected to continue as companies compete for customers, who now have smoother paths to the products and services they want.