RPA or (Robotic Process Automation) is a way to automate business processes by applying technology as a solution which is governed by business logic and structured inputs. There are many tools like Blueprism available in the market to facilitate and automate basic operations.

Those who have been workingwith Excel Spreadsheets can see this as a process that builds Macros onsteroids that can do much more than just handling Excel Spreadsheets but can do many data-oriented tasks like Data Entry, Data Clean up, merging datasets, validating various web forms etc. and free up the user’s time to do more complex tasks that require their attention. Many see RPA automate mundane rules-based business processes and others see RPA as a step towards adopting machine learning (ML) and artificial intelligence (AI) tools in the long run.

Using RPA tools, you can configure a small code (also known as a bot, a robot or an agent) that emulates process steps that you perform while doing a task. Menial tasks like capturing and interpret applications for performing a transaction, manipulating data, triggering responses and communicating with other digital systems. RPA scenarios range from something as simple as generating an automatic response to an email to deploying thousands of bots, each programmed to automate jobs in a big ERP system.

Capabilities of RPA as a tool?

RPA is a great tool to minimize human errors rather than purely seen as a tool to reduce staffing costs. Any automation discussion triggers the fears of job loss and insecurity, RPA is no different, but the primary discussion point here is Capability, not Intent. Many processes have a maker-checker format where checker take scare of the fat fingers, general human errors and in certain cases risk assessment of the transaction that is getting processed. Of course, the risk assessment is something that still needs to go through 4 eyes check but, if that check is set up at the trade or transaction initiation then rest of the transaction can be automated by RPA as a standard function.

As the formulae and automated process pick-up the values from one set to another the chance of human error goes to zero and the outcome becomes a sure shot. This in fact, makes the testing for any future changes more critical as any test case missed to be tested, then lead to an automation of the issue that is induced at the start of the process. A classic case of GIGO (Garbage-In-Garbage-Out).

The potential to save several man hours in screening the transactions/processes by human checkers makes it a lucrative option to explore RPA as a standardized automation exercise in the workflows and approval processes.

Image Credit: www.edureka.co

Another evolving use case of RPA is developing in preparing test datasets for nonproduction environment. For UAT environments, the usual exercise is to get the production data and mask the customer info or any other sensitive information and proceed, but the biggest challenge comes over when data for Test environments is to be prepared. I’ve seen QA teams struggling to build the data sets that imitate the scenarios that can come in Production systems and match the complexity, so testers can be satisfied with the code performance.

This is a huge manual exercise and a lot of things can be simplified with RPA stepping into this space. I know Glee Trees in Singapore is working in this space and have achieved some success in emulating this process. There are more firms in this area, but I haven’t yet conducted my research in this space. Will post a details report on this in the days to come.

What RPA can’t do?

Like any process or solution RPA isn’t a silver bullet and is not for all organizations, but there is a huge business case for all the financial services organizations, especially Banks that have a lot of manual processes built up with time. Forrester Research estimates that RPA software will threaten the livelihood of 230 million or more knowledge workers or approximately 9 percent of the global workforce.

Installing a bot that is foolproof and does its work extremely well takes a lot longer than originally estimated and is more complex and costly than most organizations have hoped it would be. The platforms on which bots interact often change and the necessary flexibility isn’t always configured into the bot. Moreover, a new regulation requiring minor changes to an application form could throw off months of work in the back office on a bot that’s nearing completion.

What can cause your RPA Journey to derail?

A recent Deloitte UK study came to a similar conclusion. "Only three percent of organizations have managed to scale RPA to a level of 50 or more robots”. So, it is definitely more than what meets the eye, and of course, a lot of things don't get highlighted in the hype. Development cycles and managing the bots is another long-term exercise that needs focus and planning when thinking about RPA as a solution.