RPA is a fast way of automating repetitive tasks. It can be used quickly, across system boundaries and with little coding work, and is therefore especially suitable to add business value in a short time. The control and deployment of IT, with all the frustrations that go with it, remain limited. Add Artificial Intelligence (AI) and innovation is even faster. But what is the relationship between AI and RPA? And what does that combination mean for your organization?
We have been in a transition from desktop software to SaaS for years. Everything has become an online platform and almost all work is done in the cloud via the browser. Entire organizations are 'SaaS-ed'. Employees spend their day in dozens of SaaS applications. Accounting, invoicing, Office 365, HR systems, marketing, CRM, time & attendance, etc. A lot of work requires you to neatly go through all those apps and collect the information you need. And because many apps overlap, you do things twice. Like entering a customer code or an invoice number. There's a salvation in "cutting & pasting". But is this our ideal picture of how software should work? Is that our idea of job satisfaction? Wouldn't we be able to free up a lot of time if the apps would understand us a little better? Then we would have time for the fun stuff again: the real customer contact.
In the SaaS landscape, RPA proves its worth. RPA makes working with all these cloud apps much more efficient. People have to deal with the tasks that require creative thinking and add value for the organization. Clicking back and forth between browser tabs, a robot can do it faster and better. Cutting and pasting is then no longer necessary if the robot understands us. AI applications are now also offered as SaaS. All kinds of AI cleverness, such as object recognition on photos or looking up an invoice number, are available as 'Lego bricks' and only need to be 'clicked' in your process. This allows you to quickly deploy AI in your RPA processes without having to know the technical details.
There are cognitive robots, which - like a kind of digital assistant - help employees to make decisions. There are also robots that go even further and make their own decisions. They have one thing in common: without data, they do nothing at all. A human being sees in a second whether a document is an invoice. A robot has yet to discover that. You do that by showing him a lot of invoices. But it usually quickly becomes more complicated: what, for example, is a ghost invoice? How do you spot fraud? How do you link an invoice to a transaction, a customer or a supplier? It takes a lot of work to program this knowledge into custom software. And if the layout of a document changes, you can start all over again. So we work with AI robots.
A learning AI robot uses a feedback loop to get better at what it does and correct errors. The decisions made are judged by people. That judgement is fed back to the AI robot. For example, a robot learns which document is an invoice and where on that document the invoice number, company name and other data can be found. If the layout changes, the robot will initially make errors or notify a user. Through feedback, however, he also learns to recognize the new invoice again. All this without having to change the robot's code. A robot is soon 85% accurate at the start of such a project. And that accuracy grows afterwards because the robot learns.
RPA and AI don't just change how your processes run. They change how you look at processes and automation. So the main mistake you can make is to treat them as automation tools. It's not about buying yet another SaaS package in the enterprise: RPA and AI require a different look at a strategic level. Because RPA is a low-code platform and doesn't require traditional IT projects for system integration, you can see that automation projects with RPA can suddenly be done in weeks, rather than months. Rolling out a new process so quickly is sometimes a revelation to users. The addition of AI in the processes makes RPA so powerful that it has an almost immediate strategic impact on the business. Many CIOs and CTOs have yet to realize that this revolution of 'Intelligent Automation' is underway and that they can create business value much faster by combining RPA and AI. Companies that shift the focus to processes, data and the deployment of AI immediately accelerate their digital transformation.
There is always a perception in your company of how the process is. You have somewhere a row of binders, or a SharePoint folder full of process diagrams that describe it. But is the process really like that? An RPA process generates data and shows you how the process really works. Often it is slightly different. For example, you see interesting 'edge cases' that are not in your descriptions because they occur so rarely. There comes a point when you can no longer manage all that data manually analyze. And that is also where AI comes in. This development, process mining, is the next iteration where RPA and AI technology come together. With process mining, you get automatic insight into the processes within your company and you get proactive proposals for how you can improve them. These proposals can then be immediately automated within the platform. This creates a digitization flywheel that allows your company to achieve its digital goals faster.
This is how technology shows you as a manager what your next step should be. You have exact insight into which tasks are performed in which processes. So you also know how best to grow. Do you need more people? Or do you need to invest in technology? As soon as you can make these decisions based on data and carry them out without being hampered by technical limitations, your digital transformation has only really begun. So don't wait too long and start exploring the possibilities of RPA and AI now. Every day that you wait is a day that leaves your business value behind.
I understand.
