Tech mahindra
Tech mahindra

Going Beyond Traditional RPA

Posted by: Kavuluru Rajasekhar On February 13, 2018 11:10 AM facebook linked in twitter

For those who are not so familiar with the new wave of automation, it is the next frontier of information technology that’s helping businesses evolve from manual to intelligent processes. It gives an impression that once the process is automated it can do everything by itself without any backend manoeuvring or management.

Let’s draw a clear, boarder line for expectations we can set from RPA and discuss about already available & in-progress enhancements that can fulfil the short comings of classic RPA techniques/tools.

Any set of tasks that is a part of a current day IT enabled business process work on a defined rhythm earlier carried out by humans. RPA, Machine Learning, Artificial Intelligence, IOT, BPM, Test automations are some of the technologies that we essentially deploy on the ground. Ongoing trends suggest, a majority of human intensive processes particularly in the services sector are best suited for automation using a combination of these tools.

Despite automation making a significant headway in terms of reading human context and developing human like intuition, classic RPA is confined only to a limited play. The role of RPA is currently confined to only those tasks that are mundane and standardized. However, with changing customer preferences and habits, it is safe to predict that any process with a customer-facing interface will require massive modifications and changes in the system.

Unless a certain level of elasticity/flexibility is built into the present framework of RPA solutions, it will be impossible to graduate onto PROCESS level automation in the true sense. Given the present scenario, at best it can be called as Robotic TASK Automation (RTA) in its current state.

Let’s look at the difference RPA can make for an end-to-end business process, for e.g. an order-2-cash process of a typical ecommerce transaction that originates on Amazon - this involves delivery of parcels from the point of origin to a given destination handled by DHL/UPS courier services.

Here, there are quite a few bulky tasks like physical movement of goods and core logistics that are NOT driven by IT systems predominantly. There are a few bits and pieces of these large business processes have off-late started adopting classic RPA for minor tasks like scheduling pick-ups/deliveries, back-end reconciliation processes in returns, etc., thus actually serving like RTA.

However, classic RPA can handle only structured data and also requires human interference for analysis. Further, any system level changes introduced to meet genuine business demands necessitate corresponding changes at the GUI layer of those IT applications.

Thus, it is prudent that we list out the top 3 hindrances of classic RPA:

  • Inability to handle unstructured data
  • Inability to handle unknown exceptions
  • Highly Analytical tasks

It is quite evident to us that unless additional ammunition in terms of cognitive capabilities are supplied, classic RPA in its current format would stay at the same level or left out as one of the transient phase in the ongoing digitization wave.

While there are multiple solutions emerging to address the above listed challenges, there is no clear and convincing forerunner in RPA products market as yet.

Among the leading 3 RPA product vendors namely Automation Anywhere, Blue Prism & Uipath, all 3 have come out with their version of -

  • OCR plugins to handle unstructured data. But the success rates are not convincing and results are highly dependent on multiple external factors.
  • AI and ML engines by leveraging IBM Watson/ MS cognitive services / Google AI to address unknown exceptions but solutions are still evolving (beta versions in evaluations)
  • Highly analytical tasks that certainly require decision making from variety of inputs which typically constitute a low volume of transactions but takes longer cycle times form another category that is still at bay for RPA adoptability.

To conclude, there should be no two opinions on the next generation of RPA. let me call it SPA Smart Process Automation that will be capable of automating sizable chunks of e2e business process with enhanced cognitive capabilities – a trend that is all set to become the driving force behind all automations.

About Author

Kavuluru Rajasekhar, Program Manager RPA Pre-Sales, Tech Mahindra

Raj has 24 years of experience in running business operations across multiple domains. He has deep understanding of varied automation techniques, which evolved over the time. Has started practicing and research on RPA right from early days of its entry into IT industry in 2012-13. He is currently serving as Program Manager in RPA practice in Tech Mahindra leveraging leading RPA products of the market to design best of the breed and competitive techno-commercial solutions to wide spectrum of clients across the globe.

Tags: Connected Platforms & Solutions, Run Change Grow
(*) symbol is mandatory field
* Email Address:


(*) symbol is mandatory field

Post a Comment

* First Name:
* Email Address:
Image Code
* Enter Image Code

Contact Us


For further information please write to

For further information please write to