As stated in this article, RPA’s adoption rate is quickly increasing as organizations are looking to automate more and more. In addition to the most frequently asked question, “What is RPA”, the other most common question is, “Can RPA replace an API?”
An API is an acronym for “Application Programming Interface”. Essentially, it’s a way for an application to expose methods externally that other applications can consume and use it to manipulate or pass data to that application. Sound familiar? It’s very similar to what RPA is attempting to do. There is one major difference, however. An API is a backend way to do this and RPA generally utilizes the front end. That being said, an RPA development environment like UiPath Studio, which is built on the .NET platform, can access an API natively in code or activities just like any external application can.
Leaving that aside for the moment, the question comes back to, which is better?
The answer is actually quite simple but there are some nuances. On a function for function level an API is always preferred. For example, if the API can create, edit and delete a customer record in an application quickly then there is no choice. Granted RPA can do the same task on the front end, but if the functionality exists via an API it the safest and best choice available.
Aside from this specific case, the decision to use an API and RPA gets to be murkier. One of the great strengths of an API is that it is highly defined. That is also where it begins to fall short. APIs by nature are predictable and the parameters you pass must to be in a certain format and type. Also, they are generally defined by core functions to the application. So, if there is something you need to do outside the scope of the API you are limited.
Enter in RPA.
With Computer Vision and the supporting toolset (i.e. .NET) you have the ability to utilize an API and use the front end of the application to fill in the gaps. If an API doesn’t exist, no problem, RPA can do it. RPA is application independent, if the robot can see it, it can interact with whatever interface it needs to connect to.
So to address the original question, which is better? The answer is “both”. I will say RPA can fill a lot of gaps whether they are narrow or wide and should be considered next time you are not able to get the data needed to perform the appropriate function in an application via an API.
Peter S Camp is the CTO and Founder of CampTek Software. He has been developing RPA Applications for over 15 years. For further questions, discussion or inquiry about CampTek Software Services, contact info@campteksoftware.com.
