An Automated Future?
There are few tech topics as popular today as automation. Automation isn’t a new tech capability of course, but it has crossed over from just mechanical automation (robotics), to include software automation like robotic process automation (RPA). Without writing a complete history of process automation, I’d be remiss to make it sound like using software to automate processes is new either. There’s an older category of software, business process management (BPM) that grew out of the effort to modernize business processes in the 1990’s, called business process reengineering. Reengineering was a popular topic among enterprise software providers, large system integrators and businesses in general as they moved to implement client - server based (versus mainframe based) enterprise systems. A lot of that was driven by the Y2K scare of course, but also efforts to “modernize” business systems and underlying technology. The reengineering was code for “the software is expensive to customize and customization makes it very hard to upgrade (you can interpret that as the software is very rigid and inflexible if you like), we made it with the ‘best practice’ processes built in and you should change your business to match them, not the other way around”. Out of that the BPM software grew up to try and support / make easier all this process reengineering…it didn’t, by the way. BPM overall was also difficult to implement and ended up very inflexible in many cases as well. The term, once referring to the software tools, has moved beyond that to focus on the discipline of using various methods to discover, model, analyze, measure, improve, optimize, and automate business processes. Another capability that emerged around the same time was embedded “workflow”, or the capability to build a process flow inside an application to automate simple tasks like approvals or forwarding a document.
Okay, enough history, suffice it to say that BPA, workflow and other software automation technologies are light years ahead of what was in use in the 1990’s and early 2000’s. Many (maybe most) enterprise systems include some sort of a workflow engine in its platform that can accomplish simple to medium complexity automation of processes generally inside the bounds of an application or maybe across a suite / cloud. This capability is particularly useful in event driven activities where a trigger action can cause the software to perform some subsequent set of activities based on the type of and timing of the event trigger. Workflow, for example, is used in customer journey mapping. The map / journey that is built and deployed is connected to the customer record, marketing automation (MA) system, sales force automation (SFA) system, and maybe customer service and other customer facing systems. The “event” could be any prospect action that sets off a sequence of events, for example the prospect downloads a report from the provider website that is behind a login wall. In the process the prospect provides an email address and permission to email them in the future. This action triggers an email drip campaign that provides additional educational material related to the subject of the report and the providers solutions on some regular schedule designed to start a conversation and ultimately create some relationship.
Beyond workflow, which has evolved into a very powerful platform, other forms of software based automation are available. BRM tools seem to be morphing into low code / no code development platforms, which provide the capability to extend and add new applications to many types of enterprise solutions. No code development platforms are particularly interesting as a way to extend lightweight development capabilities to end users, which frees up IT to focus on projects that require deeper tech expertise. Simple software bots emerged over the past ten or so years ago and at first were used to automate a variety of fairly simple tasks to make employees more productive by eliminating human intervention in routine tasks. Bots capabilities have grown and intelligent bots, as well as RPA dramatically expand the tasks that can be automated.
RPA
RPA leverages software bots to automate tasks inside applications that would usually be done by humans. That’s the difference between RPA and workflow; RPA automates human tasks that would be manual and workflow automates application activities based on event triggers. RPA automation focuses on tedious, routine, repetitive tasks inside an application, or between several applications. The two are similar and both provide tools to configure the automation based on user needs. In RPA the platform is also no code, and usually provides a form of visual tools to “drag and drop” building the automation or record the actual human actions and make them repeatable. The automation can function independently of humans or work in concert with human interventions and actions.
Automation and Change
Automation can increase productivity, improve customer and employee experiences and generally provide a lot of value to the business. In the future the capabilities can only get more powerful and grow in scope. RPA is one of the hottest automation tools over the past 5 or so years. Process automation across the business is touching nearly every function. Just applying RPA can:
Increase productivity and efficiency by applying automation to the tasks that it performs well, and freeing humans to focus on tasks that require human intervention
Deliver an improved customer experience (CX) using intelligent bots that can both handle simple customer questions and requests, and accurately direct (quickly) the customer to the human that can deal with their issues
Improve the accuracy of repetitive tasks that are generally error prone for human execution
Reduce disruption in businesses with legacy systems by providing a path to automation that doesn’t require “rip and replace” approaches
Improve business processes by identifying gaps from the data generated by RPA bots
Provide reserve capacity to absorb irregular tasks volume
As you can see, automating processes using RPA and other automation technologies has the potential to provide a great deal of benefit to the business. Process automation does come with some business risks as well. In other words, automating what’s broken doesn’t solve your problems. In any automation project it’s critical to examine the underlying process and determine if you need to update process and/or supporting software systems to get the outcome you desire.
Intelligent Automation
As software automation capabilities continue to evolve more systems are incorporating intelligence, or the capability to independently learn, interpret and respond over time using data and ML, NLP, deep learning, etc. RPA is automation that is generally applied to “rule-based”, process oriented, repetitive tasks; simulating human tasks actions, but not applying intelligence. Intelligent automation (IA) adds intelligence to the process with the ability to interpret data, make inferences and conclusions on based on the data and take subsequent actions. IA then, is built on RPA, but adds many capabilities beyond simple task completion and moves beyond the rules-based approach. A simple example is support chatbots. First generation, RPA like chatbots can only interact inside the limits of its pre programmed responses and rules, pushing the customer to a predefined webpage with an answer / FAQ, the correct support agent or some other predefined response. “Smart” chatbots can interpret the customer during the interaction and infer the proper reaction. The IA adds the capability to function in an unstructured data environment, moving outside a logic tree set of responses. this gives smart chatbots the capability to handle exceptions and work outside the “script”.
Automation is applied to all sorts of systems and business problems. RPA can provide a great deal of value relieving humans from the routine process actions and increasing productivity and outcomes by giving the humans the space to handle exception and higher level tasks. IA goes much further than simple RPA, and add the capabilities to work outside the program framework, inferring, predicting and taking actions based on the unstructured. It also learns and improves its performance over time. IA then, is essentially the next generation RPA, supercharged by applying AI technologies.
