Robotic Process Automation (RPA) applications are ideally suited to executing routine and repetitive administrative tasks traditionally performed by humans. These tasks include “swivel chair” functions that move data from one system to another, processing routine documents and verifying account information. As such, the assumption is that RPA primarily involves relatively low-skilled and low-paid workers.
In certain instances, however, RPA tools can have a significant impact on positions commanding salaries well into the six figures. Consider commodity traders – they are well compensated for their expertise and ability to analyze data and make decisions on how a variety of global economic and political factors might affect the value of crops, fuels or currencies. The decisions they make help determine whether financial institutions earn or lose millions.
Esoteric, specialized and valuable as a commodity trader’s skill set may be, the reality is that these highly paid professionals can spend three to four hours a day moving data around from one application to another – doing precisely the type of mindless, mundane tasks that RPA is designed to do better, faster and cheaper than people.
So what does this mean for an RPA business case for a commodity trading desk? It’s one thing to automate 30 percent of the job of a healthcare claims processing agent who makes $15 an hour, and quite another to take out 30 percent of a job that pays ten times that. Another consideration: how would an additional 30 percent of bandwidth impact how commodity desk managers gauge the productivity of their traders?
What’s especially intriguing (and especially scary, if you’re a commodity trader) is the idea that RPA tools, coupled with cognitive applications, could potentially replicate significant portions of a commodity trader’s expertise. Consider: trading decisions are based to a large extent on historical patterns and cause/effect linkages – for example, how changes in the price of energy impact the price of crops. A trader applies knowledge and experience of those historical patterns to make those associations. By leveraging pattern recognition capabilities and if/then logic rules, emerging cognitive tools are increasingly able to replicate that type of sophisticated knowledge work.
The lesson here is that enterprises seeking to leverage intelligent automation to its fullest should look beyond the obvious targets of functions characterized by drudgery and boring, repetitive tasks. The potential of smart tools to impact high-value, high-skilled jobs suggests we’ve only scratched the surface of the art of the possible. And those of us who are confident that our jobs are immune from the robotic onslaught may want to reconsider.