Career Update: I am now an RPA Developer! It's funny how I have a developer title in my job description now without having a CS or programming degree haha. It's been an unexpected turn of events but nonetheless a pleasant one. I'm still aiming to learn ML and predictive analytics but let's just say this is a 'canon event' in order for us to build a solid ML and Predictive solution. So enough of that and let's dive right into this three-letter field.
What is RPA?
RPA is short for Robotic Process Automation. As the name suggests, it is a method of automating business processes with the help of a robot or tool that can replicate human actions or transactions. This still covers the ETL process to digest the data but instead of a call to action as an output, It immediately does the needed action. Instead of insights, it gives you or your colleagues legroom to engage in other areas of your work and further your capability to provide value. It is primarily applied to processes that are either manual, repetitive, follow a set of definable conditions or a mixture of all those things. So less daily zombie clicks, and more checklist ticks!
Use Case of RPA solutions:
Aside from the automation aspect, RPA solutions are also flexible solutions crossing boundaries and platforms. Usually, it lets you apply the needed action from one platform to another. Like from ETL tool to a CRM, CRM to CRM, CRM to ETL tool to CRM, Basically any combination as long it's needed to complete the designated action.
A specific example could be SAP Transaction and Management Automation, Salesforce Case/Item automation, or anything that you think needs automation and requires a great deal of back and forth and repetition on your side. I cannot dive deeper into examples because I might be flagged for that haha. Just remember these three things to spot an opportunity for RPA development: It requires manual work(Excel work), it is repetitive, and It can be guided by a set of definable conditions.
Platforms used for RPA:
I was shocked to know that there's a whole array of tools that are used in RPA development. Also from what I currently know there are two types of RPA approach. First, there is the backend RPA approach, just like in Data analytics we touch on the data or code that is stored within the platform, so all the changes and procedures are done on the data or the code directly. Next is the Front-end approach, basically the bot acts as the human that interacts with the interface. It does the required clicks and responses just like when a human does it.
Sample platforms are:
a. Blue prism - this tool is somewhat like Alteryx wherein it utilizes the concept of flow chart to do the step-by-step process. But unlike Alteryx this tool is script or code-intensive in terms of input to do a certain action. However, they also utilize macro-type concepts to make widely used processes readily available to lessen redundant work.
b. Uipath- This tool I don't really know but some experts say it works well with Alteryx.
c. Alteryx- Yes Alteryx, Our beloved Alteryx can be used for RPA(The main reason I was hired haha) Although this is limited to available connectors for each platform. Along with current expertise in Alteryx, each connector may differ in syntax so there is the point of learning and discovery when we use Alteryx
Is getting into RPA Development easy as a Data analyst? Yes! Unexpectedly It was easy. maybe because I utilized Alteryx for RPA development and it was just a matter of learning that specific connector and incorporating it into the workflow. Also as a Data Analyst, the similarities outweighs the difference so it was a somewhat smooth transition.
Similarities:
a. ETL Process - In RPA ETL is a crucial part of making it possible to clearly instruct the Robot on Datapoints needed for the Business process. Like listing all the Orders for the process and making sure it qualifies for the condition set.
b. Focus on Automation - Back in the days of dashboard making we emphasized automating the data so we see data in real-time and lessen manual work for finding insight, the same applies to RPA. We make sure that the tool can run on its own without constant supervision and it is set to run any given schedule it follows
c. Agile Approach - Scrum is the main methodology used and is surely perfect for both RPA and DA. For the same reasons that we are able to segment work and produce MVPs on a short amount of time or sprints
Differences:
a. Platform Dependency - As mentioned above RPA is used primarily on different platforms, this makes it diverse in terms of how we are able to make the solution work. Different scripting/structure/syntax is apparent in this line of work, so it is heavy on learning different parts of the platform first before we are able to command our Bot on what to do and how to do it properly
Does predictive analytics and ML go well with RPA?
Surely, I see RPA used in two ways. First is a dynamic way to prepare and get data. For Example, you need an input that is sourced from a given transaction, instead of waiting for people to manually process the transaction you utilize RPA to make the process as seamless as possible with less downtime or being prone to manual mistakes.
The second use is as a hand for your AI. So what will you do with the insight you got from your AI? Let RPA handle that and make the process seamless and efficient. Like predicting inventory by use of AI then let RPA place the orders for you. Just make sure you put in an approval step because you might encounter a bug and instead of buying 100 inventory, you buy 10,000 haha Surely you cant scold your computer :D
Is RPA an End-Game Solution?
I view RPA as a bridge, a bridge that connects your business from a manual process to a long-term automated sustainable solution. It is a needed step to achieve business automation excellence. Because it provides the solution to the now problem. Later on, it becomes the foundation for solution incorporation in the current platforms. It might not be the End Game but it sure is an essential step forward the End Game efficiency we want to achieve.
コメント