There’s a renaissance happening in organizations today. Process automation is now ‘in vogue’ again. It’s no doubt that robotic process intelligence (RPA) and artificial intelligence (AI) are driving this renaissance, helping to transform the enterprise. But what that transformation looks like is another question entirely. Even while the goal remains consistent—to elevate the customer journey and improve end-to-end business outcomes—the path can vary dramatically, depending on the technology used as a foundation for transformation. RPA, featuring smart software robots, has gained considerable ground on this front, based on its ability to quickly improve productivity, reduce costs, and drive a competitive advantage.
But confusion about RPA’s relationship with AI is still a factor: which use cases should use AI, where to start, and which technologies work together or complement one another? Bandwagon messaging and overhype has not helped answer questions, and to the contrary, is feeding misconceptions about how each technology works and which are ideally suited for various business processing operations. As AI becomes more widely embraced and proven with real-world deployments, these messages will become clearer—but smart CIOs can take an early lead by defining tangible problems to solve and worrying less about the technology terms themselves. Insight from Gartner supports this position, suggesting that CIOs are best-served with the simplest approach that will do the job, rather than buying into potentially difficult and costly, albeit cutting-edge, AI techniques.
Diving into RPA and AI
AI and RPA are naturally complementary. But understanding the RPA/AI relationship is made more complex by the spectrum of AI technologies from which to choose. AI has become an umbrella term, covering machine learning, neural networks, natural language processing, predictive analytics, cognitive automation, and more. Each of these AI technologies has its core strength and is typically designed to handle specific challenges, rather than general purpose use for solving a wide variety of business problems. So where do you begin? First, start with the business problems to solve, defining requirements to address the most imperative aspects. Next, identify ‘cognitive moments’ to apply AI where it adds value improving particular, non-ambiguous cognitive components of the overall solution. Don’t try to boil the ocean. As an example, AI can be excellent in augmenting employee customer service activities. However, initial customer-facing AI implementations that sought to completely replace the human element typically failed to meet expectations. These deployments actually resulted in a poor customer experience, demonstrating that AI today can be complex to implement. More effective uses are tightly defined, non-ambiguous uses, or cognitive moments. This is why many organizations are now taking an AI+RPA augmentation strategy, instead of completely replacing the human element.
CIOs can get early wins and fast time to value by not going overboard and trying to use too much advanced learning technology up front. Scenarios in logistics, accounting, and banking tell the story, demonstrating RPA’s comparative maturity and its role in enabling compliant, repeatable processes.
RPA in Action: Transforming Logistics
“We’re seeing a 95% reduction in manual effort from our customer service representatives. Our RPA platform handles what our users were doing, automatically, more consistently and more accurately,” said Darren Klaum, Director of Business Systems for PITT OHIO, a third-party logistics (3PL) provider.