Process optimization is a discipline as old as civilization itself, and until recently, it was the domain of humans alone. But intelligent automation is the next frontier, and it’s poised to provide countless benefits to organizations aiming for a strong ROI.
Thanks to the growing capabilities of artificial intelligence and its subcategories of machine learning and deep learning, computers can not only automate repetitive processes but also recommend improvements by scouring quantities of data too vast for humans to comprehend.
This type of automation has the potential to deliver breakthrough ROI by enabling organizations to fundamentally redesign the way they work.
Technology grounded in RPA
End-to-end intelligent automation builds upon robotic process automation (RPA), a fast-growing category of software that automates the activities of knowledge workers. Gartner estimates that the worldwide robotic process automation software market will be worth nearly $2 billion this year and will continue to grow at double-digit annual rates through at least 2024.
Without intelligent automation, many knowledge workers spend a lot of their time on routine tasks that can’t be automated because the technology to do so doesn’t exist. Much of this low-value work involves data entry, such as rekeying numbers from a legacy accounting application into a new enterprise resource planning system or transferring data on printed forms into databases.
Take a use case like processing a mortgage application, for example. Applicants typically submit numerous bank and investment account statements as well as documents filled out by hand. This information must be entered into software that scores loan applications, which can be a manually intensive and error-prone process.
One technology organizations use to try to address this type of issue is RPA. With RPA, businesses can save time on tasks and enjoy greater ROI, as RPA software agents called bots can be trained to automate much of the detail work by using optical character recognition to, for example, translate printed input values into the requisite input forms. These bots can also flag exceptions that require human attention, a process sometimes called “swivel chair integration.”
Moving beyond individual tasks with intelligent automation
RPA has a lot of value, but it tends to be confined to automating individual tasks rather than end-to-end processes. RPA can also have the unintended side effect of paving cowpaths, as old and inefficient processes can become hard-coded and, as a result, more difficult to change. That’s where intelligent automation comes in.
By using a combination of AI, machine learning, and end-to-end process automation the enterprise can create intelligent workflows capable of learning, finding opportunities for improvement — and then implementing them.
Let’s take our loan processing scenario as an example. Suppose the process requires analysts to examine 15 different sources of financial information as they review a typical application. Machine learning algorithms can crunch millions of prior applications to determine, for example, that just eight sources of information correlate highly with the loan approval. It can then recommend that lenders reduce the number of documents they need to examine and the time needed to decide.
Machine learning can also pinpoint steps in processes that may not be necessary at all. Perhaps the number of people required to gather documents can be reduced, or some records may be able to be retrieved automatically by obtaining the applicants’ permission to access electronic accounts. These efficiency-enhancing opportunities often escape the attention of human operators because they are too accustomed to existing procedures or lack the information to envision alternatives. When full end-to-end systems deploy intelligent automation, workflows become more productive and businesses glean higher value from their processes.
The low-code equation
Once domain experts are equipped with process improvement insights, they can use low-code programming tools to redesign their work. According to Forbes Technology Council, “the introduction of low-code/no-code automation resolves these issues, breaking down barriers and empowering teams of any skill level to manage IT operations seamlessly.”
Low-code tools are important because process optimization is rarely a “one-and-done” proposition. By their nature, machine learning algorithms become more effective over time, which makes optimization a continuous journey.
Low-code tools enable anyone involved in the business process to model workflows and modify them easily. Low-code platforms can also integrate services such as image recognition and voice response via application program interfaces, making these services accessible to developers of any skillset.
Collaboration is the key to realizing the potential of intelligent automation. Process designers can work across departments to co-create processes using simple drag-and-drop and visual tools. Professional developers and data scientists can look at the underlying logic and data models to create enterprise-grade applications where needed.
Reaping the benefits of intelligent automation
Process optimization will become even more important as self-service assumes a more prominent role in customer engagement. In fact, Gartner has predicted that 85% of customer service interactions will begin with a self-service component by 2022, up from 48% in 2019.
The ability to capture customer data digitally at the point of origin presents opportunities to kick off even greater improvements. That’s how the Municipality of Dubai eliminated paper almost entirely from its operations, in the process creating a single external-facing portal for its more than three million residents. Instead of visiting multiple departments to get the signatures required to open a business, residents can now interact with government agencies through a single portal. Because the resulting data is captured digitally, it can be fed into machine learning models to identify continuous process improvements as well.
As these advancements in intelligent automation continue to gain traction, remember that such improvements don’t remove the need for humans; they just enable them to spend more time doing work that only they can do.