A lot has already been written about RPA and AI. But one thing is certain: they are here to stay. Because make no mistake, automation is no longer a thing of science fiction. The ability of RPA and AI to increase productivity, efficiency, and customer satisfaction ensures that every company in every sector can now reap the benefits.
However, the fact that automation is dropping its ‘technical’ image does not mean that all the jargon involved is suddenly understandable to the uninitiated. Since we’ve noticed that there’s still some uncertainty about the actual difference between RPA and AI, and that they’re often even lumped together, we’ll explain that difference in more detail for you in this long read.
We’ll also explain how RPA and AI can work together. So that it doesn’t get too theoretical, we’ll do this using two simple examples within finance and HR – two domains in which RPA and AI work quite well.
definition of RPA and AI.
what is robotic process automation (RPA)?
RPA works via software robots that take repetitive and less valuable work off your hands. Those bots are real doers: they follow strict rules and carry out their tasks flawlessly. It is important to remember that RPA requires structured input to produce good results.
Software bots can be developed and implemented fairly easily and fairly cheaply. Because of this, Robotic Process Automation provides you with good results quickly. And that speed often turns out to be crucial when it comes to optimising your business processes and achieving your objectives.
RPA supports processes in your company such as:
- Application logins
- Copying and pasting data
- Filling in forms
- Opening emails and attachments
what is artificial intelligence (AI)?
AI usually requires more implementation and follow-up work than RPA. AI refers to the ability of computer systems to mimic the human brain. They are able to make judgements, make decisions, and learn. Plus, AI can work with unstructured input, as well, given that the algorithms are capable of carrying out complex reasoning calculations. In other words, AI algorithms are thinkers.
You can achieve complex objectives with AI. You probably already feel that developing and setting up such an automation process requires quite a bit of effort, certainly in comparison with RPA. However, this effort is not in the development work, but rather mainly in learning and testing. Of course, you’ll see a high return on these investments in the long run.
Some of the possible applications of AI:
- Understanding documents
- Converting speech to text
what about the combination of AI and RPA?
Although AI and RPA are clearly different, you can’t see them as totally separate entities. After all, if they join forces, they can take automation, and thus process improvement, to the next level.
But how do you create that combination?
We could write lengthy articles about which steps you should take when you decide to automate your processes, but that takes us off on a long tangent… And isn’t our goal to remove any confusion, rather than creating more questions? ;-)
If we specifically take into account the scope of this article, then – very simply put – you proceed as follows:
- First, it is best to map out the simple processes.
- Then you check which of those processes are time-sensitive and error-prone, require a lot of manpower, and have an impact on other processes and systems.
- Such processes can then be made more efficient using RPA.
- If that basis is there, you can add AI to workflows that are too complex for RPA alone.
This way, you scale up your automation solution gradually and lower the threshold for integration with AI (and more).
example 1: processing invoices.
Processing invoices is such a process that’s fairly easy to map out (step 1 from above):
- You receive invoices by email.
- You put those invoices in a folder.
- You extract the relevant data from them.
- You add the invoices to your accounting program.
Is the process time-sensitive and error-prone (step 2)? Yes – it must be done on a regular basis and within a certain timeframe. And if it’s a manual process, it’s also error-prone. There is quite a bit of work creeping in and the processing can, if it’s not done on time or correctly, have an influence on the rest of the process flow.
Let’s see what RPA can do then (step 3). Software bots can be used to:
- Retrieve the emails
- Download the invoices to the right location
- Copy and paste the invoices into the accounting software
Robotic Process Automation has already yielded a number of quick wins here. But then, of course, we’re not quite done yet. AI can further optimise the process (step 4).
Not every invoice looks the same: not all information is in the same place and not every invoice has the same number of rules. This means that we can place invoices under the heading of unstructured data. And as you remember from our definition above, RPA cannot handle such data. Therefore, AI is required in this case to intelligently read the invoices and recognise and process data such as invoice number, due date, and amount to be paid.
example 2: the recruitment process.
You can do a lot to enrich your recruitment process with RPA and AI – especially in the case of AI. You could analyse facial expressions during interviews or automatically perform detailed background checks. Such applications are not covered in this example. Here, we focus on how AI and RPA can lead to time savings, error reduction, and greater efficiency.
Recruiting a new colleague entails a lot of different tasks:
- Filling in forms
- Screening CVs
- Adding data to the databases
- Providing information
- Scheduling training courses
Should you be worried that AI and RPA are partly taking the human aspect out of the hiring process? Not really. Of course, it’s always a good idea for a company to list the pros and cons, and to determine how far you want to go with it.
However, more often than not, your HR employees are very busy. Because of this, they might not always be able to pay as much attention to personal relationships with applicants or new colleagues as they would like. If you automate certain burdensome tasks, you can gain extra time that can be put back into the welcoming, human aspects of the job. And the experience for applicants and new colleagues will only improve by avoiding mistakes and sloppiness!
But which tasks are can be done using RPA and AI? Below is a small selection of the possibilities.
questions about routine issues.
Potential candidates often have questions about the job content, the recruitment process, or the terms and conditions of employment. New colleagues are already asking about leave requests or working remotely. A chatbot not only contributes to conversational recruitment, but can also help to instil a strong employer brand feel among new colleagues.
If the intervention of a human colleague is required, the question immediately reaches the right person. At the same time, the communication on routine matters is kept to a minimum. Handy!
initial screening of CVs.
Manually screening CVs takes a lot of time – especially if you already know that there are many candidates in that stack of CVs that do not meet the job requirements. Thanks to AI, it is possible to have all incoming resumes screened automatically. Based on certain characteristics and terms, a judgement is made, after which the CVs can be ranked on the basis of suitability. Scheduling interviews with the best candidates? This can also be done automatically for you.
We’ve already mentioned that a chatbot is perfectly capable of answering simple questions from a new colleague. But RPA and AI can also make other aspects of the onboarding process significantly more efficient.
Onboarding often equals paperwork. Thanks to RPA and AI, it is possible, for example, to automatically request standard information from the new colleague, such as identity card details and degrees. And a number of data can already be retrieved from the CV or database and copied to digital forms, so that the same information does not have to be filled in every time.
If you add a new colleague to the database, you can add an automatic flow to their profile that makes rules-based decisions. For example, your automation solution knows in which department the new colleague has started and can then send out invitations on its own for the proper training courses or welcome sessions. Or depending on the position, the employee is automatically granted access to certain business systems and data.
Are you interested?
You will notice that the application possibilities are almost endless. It is important to properly map out your current processes, determine your objectives, and then look for the right solution.