24/5/2023
AI and RPA, a golden combination
As you have noticed without a doubt, AI is currently a hot topic of conversation. The lightning-fast development of this technology is already bearing fruit in many areas, and for us it offers many new possibilities. We recently applied AI for the first time in an RPA robot running in production and we are impressed by the level of accuracy that can already be achieved with it. The influence of AI will only increase, is the general tendency, so we are happy to share our views on this.
Current application of AI in RPA robots
As mentioned, we applied AI for the first time in an RPA robot, specifically for recognizing and storing certain data from supplied documents. We then use this data as input for the robot. Without AI, we would have to write a lengthy script and specify exactly where in the text the required information is located. With around 40 variables, this is time-consuming and maintenance-intensive.
The sometimes large differences in structure between different documents do not make this process easier. Using AI, after a number of training documents are delivered and processed, we can deliver new documents and the location of the requested information is automatically recognized. The best thing about this technology is that it continuously improves itself. We see that with a small number of training documents the accuracy is already almost perfect, so you can imagine that this process works so efficiently that it soon has nothing left to be desired. We are very excited about this and are already thinking about other possibilities to make processes even better using AI.
Other possible applications of AI
A more general possible application of AI that we are excited about is the possibility of systematically applying the technology for a client who would like to implement it in a particular process. For instance automatically generating text that would normally be written manually. This could potentially save a lot of time (and costs). Specifically, then, in this example, certain keywords could be supplied, along with some examples of what the text should look like. This would allow AI to generate a new text with the desired structure incorporating the keywords. You can, of course, provide this to an AI service yourself, but then you still have manual work inputting the variables and copy pasting the results back to your own applications. We can fully automate this, even as part of a larger RPA process. This example is just one of many conceivable applications of AI in combination with RPA.
AI and VionA
It goes without saying that we at VionA keep a close eye on the developments regarding AI. If you want to apply AI but are not sure how or if you need help in any other way, please do not hesitate to contact us. We will be happy to answer your questions, discuss your possible doubts and see if the application can be productive for your organization.