AIOps Vs RPA

Race for digital transformation

I first heard the term RPA a few years back and it immediately sparked my interest. The idea that business processes could be automated and completed within seconds by a robot sounded extremely futuristic, but equally engaging to me

Last year, I heard of AIOps for the first time in a report about the evolution of IT from Gartner which got me really excited. Artificial intelligence, machine learning and big data are subjects of big interest for current IT leaders and they are the backbone of AIOps.

Now with both RPA and AIOps in the IT landscape I see a lot of confusion around them. Are they the same thing? If not, how are they different? Can they be used together in your infrastructure?

A while back, I wrote about what AIOps and its role is so I will not go into much detail about it again besides quoting Gartner: “systems that combine big data and AI or machine learning functionality to enhance and partially replace a broad range of IT operations processes and tasks, including availability and performance monitoring, event correlation and analysis, IT service management, and automation.”

RPA stands for Robotic Process Automation and it focuses on automating business processes based on logical and repetitive steps which are often encountered in business applications such as CRM and ERP.

At a first glance, one main difference between the two is that they address different segments of IT. RPA focuses on business applications while AIOps addresses day-to-day IT operations. RPA’s role is to smooth out business processes such as registering a lead in CRM while AIOps is tasked with assuring that the infrastructure is optimized and running at full capacity.

AIOps technologies gather data from within your infrastructure (servers, desktops, network, apps etc.), apply machine learning to it and act upon the results through automated processes. As such it is both reactive and proactive in the sense that it reacts to machine learning detected anomalies from within your infrastructure (if a server goes down it will automatically send a notification and open a ticket) but it can also predict what can happen in the future and apply a fix before the issue arises.

RPA is a set of predefined steps that can only go in one direction. It takes tasks which typically require no creative or problem-solving skills and completes them based on a certain logic and input. One example of RPA is creating a new contact in CRM based on message from a website form. The input are the field values from the website form which is processed and added to the appropriate fields in CRM. Once each field that defines a contact is filled the RPA reaches the end of the process and stops.

Another big difference between RPA and AIOPS is in how they are deployed and configured.

RPA usually comes in the form of an application or small process running on your computers. The steps taken by the robot to complete its tasks are defined beforehand by a human based on different variables such as field values or a selection from a drop-down menu (needless to say RPA can also complete free-form text if needed). Once the steps are defined, they are fed to the robot and the task is now automated. Should the need to automate another task arise, the steps for that task need to be defined and fed to the robot same as before.

AIOps is a much more complex solution which needs more than one technology in order to properly work properly. A big data solution needs to be in place to capture and store the data gathered from the infrastructure. This data is then processed, analyzed and acted upon through various automated tasks. The main difference from RPA is that not every step of the process needs to be defined. We define machine learning jobs to look for anomalies and patterns and what we want to happen when they are identified.

Two sides of the same coin

Artificial intelligence and automation have come a long way in the past few years and we are now in a full sprint for digital transformation.

I believe that AIOps and RPA address digital transformation in different, but complementary ways. Both solutions should be considered if a business truly desires to go on the path of digitalization.

AIOps and RPA are both focused on reducing costs, optimizing processes (business and IT), saving time and increasing productivity. All of these are reasons enough to implement either one or, preferably, both