In our most recent AITP meeting, Mary Elizabeth Hooper, Director IT – Innovation, Robotics (RPA), Customer View, Adoption and Service Measurement/Reporting at Synovus helped demystify RPA.
In a nutshell, RPA is like a macro. Many of us have used or built macros in our careers; oftentimes through tools like Microsoft Excel. Similar to the Macros of years past, core RPA technology is not thinking, is not learning, it is only “doing”. Core RPA tools follow a set of rules to perform a sequence of automated actions with the goal of acting like a human. There are certain forms of RPA that are more advanced, automatically enhancing rules engines based off of “learning” from the continuous executions of actions and observations of outcomes. A good example are many chatbots; chatbots, while appearing fascinatingly intelligent, are often simply guided by a set of rules initially programmed by a human or adjusted by an algorithm.
At a certain point, when the technology begins to “reason”, navigating the “gray-area” in between rules, it transcends from being an RPA solution to a Cognitive Automation solution. These solutions support many advanced capabilities including natural language processing. Google Home, Alexa, and Watson are all examples of Cognitive Automation solutions.
While this makes sense, what is the value to a business? Mary shared her perspective garnered from her experience at Synovus:
Large businesses, especially those that grow through mergers and acquisitions, are constantly faced with the challenge of wrangling their data. Oftentimes, key financial and company data lives in different forms in different systems; many of these systems are legacy without effective mechanisms (like APIs) to access that data. In this state, business analyst resources have to invest significant time to manually query, access, and re-key data to create the necessary aggregate reports for the business. To combat this manual requirement, the IT teams for these large businesses are often asked to undertake the difficult task of integrating these systems through a complex programming effort. So how does RPA help?
RPA allows those manual business analyst tasks to be automated through a series of steps and rules. Query a database? Copy a record? Enter it into a spreadsheet? All automated into the RPA equivalent of a macro. The end result is an automated workaround for having to deal with legacy systems, saving the time and expense of having human resources perform those tasks manually.
In a perfect world, where system effectively “talk” using standard open APIs, the need for basic RPA would fade. However, we are a long way away from that utopian view. As long as businesses are continuing to rely on disparate, siloed systems of record, RPA will continue to be a valuable tool for navigating the chaos.