A wide range of new technologies and new buzzwords have been jockeying for the attention of process owners across industries over the past year, with robotic process automation (RPA), machine learning and artificial intelligence capturing headlines and creating new buzz. Sorting out what these new technologies can offer and what practical use cases they can be applied to, can be confusing and challenging. Artsyl Technologies, a leader in intelligent process automation for back office processes like vendor invoice and customer sales order automation, has spent the past year on the ERP and ECM market conference circuit, translating the hype into reality for middle market technology customers.
Each of the technologies described above offer a distinctive value proposition compared to traditional approaches to process automation. They ALL set themselves apart from past generations of automation technology by offering greater flexibility, more out of the box functionality and more user-centric (as opposed to IT/code-centric) approaches to automation.
Broadly speaking, what that means in practical terms, is that today’s process automation tools can be set up quickly, as less cost, and can be configured or fine-tuned by an end user or process user, with less dependence on IT. This is huge for a number of reasons. One of the big reasons is that with a lower total cost of ownership, those persistent manual tasks that companies and employees used to tolerate because the solutions were too costly or complex, are now new suspects for automation.
Take robotic process automation, the simplest form of automation on the spectrum of modern automation technologies. One astute ERP system VAR, when exposed to RPA, described it as “macros on steroids.” This is apt, in part because RPA tools can record user activity/interaction with an interface, and duplicate those steps repeatedly. The key is that it is relatively easy for users to apply RPA, without complex coding. As a result, routing, standard, repeatable tasks can be off-loaded from human operators and handled by robots. Just like with macros, only using a standardized, easy-to-use toolkit.
The problem with RPA is that it is otherwise a “dumb” technology that isn’t designed to handle process exceptions well. That is a big problem—one that has proven to be the downfall of more complex technologies that allowed for modification by modifying business rules/logic to accommodate exceptions, but as a significant investment of time/effort.
This is the cue for machine learning, AI and RPA to come together to deliver the kind of solution that works in the real world. Machine learning is one dimension of AI. It implies a system or tool that can adapt and learn through experience or through input, so it can adapt to exceptions and new conditions without adding rules or inserting code. Artificial Intelligence takes advantage of machine learning, plus a bunch of other technologies to achieve a more probabilistic approach to problem solving that does a better job than humans of predicting outcomes.
So, what about automating routine processes, where there are definitive outcomes in a way that is flexible and adaptable? That’s where intelligent process automation (IPA) comes in. IPA starts with the basic framework of RPA systems, so it can repeat recorded or configured instructions. BUT, when it encounters an error, exception or situation is can’t handle, it knows to turn to the human process owner. A process owner then steps in, makes corrections manually, and records those actions. The IPA tool then takes that recording and adapts its algorithm so it can handle the issue without help from a human next time. Which means that it is applying machine learning.
Today, intelligent process automation platforms like docAlpha from Artsyl Technologies can help companies automate 80% of common business processes like vendor invoice or customer sales order handling out of the box. The system continues to adapt and learn from interactions with human operators to adapt and evolve to handle close to 100% of scenarios and use cases related to handling documents, “reading”/extracting data, matching, routing and transaction entry.
Automation of document handling and data entry means that companies have the data they need to apply to business rules for approval routing and coding. It also means better visibility to general ledger accruals and better control of cash flow. Essentially, with the data and document problems solved, the process problems solve themselves.
Today, delivering on the promise of new technologies like AI and machine learning can be practical, achievable and affordable enough to produce tangible results and ROI quickly.
To turn your vision into action, visit the Artsyl Web site and explore our resources page. Then, reach out to your Artsyl account representative to chart a course towards your digital future.