Hyper-automation in business processes and financial services deliver significant value as technologies improve, Vuram director Balaji Sridhar said.
Sridhar defined hyper-automation as combining complementary technologies that allow for end-to-end process automation.
Before the onset of the COVID-19 pandemic, companies engaged in hyper-automation, but mostly on its edges, by addressing components of more extensive processes.
More companies are expanding their vision of the concept to entire, end-to-end processes in our more remote world.
Companies should be prepared for a continuous learning circle when engaging in hyper-automation, Sridhar said.
Once a procedure is automated, they often learn that the system can smartly do some steps. They can then build rules once they understand customer sentiment or add some machine learning.
“The next component of hyper-automation is where we are teaching our systems to become smarter and making them capable of taking their own decisions, to some extent, with minimal human intervention,” Sridhar said.
Mining, interpreting data
Concurrently, the systems are mining and interpreting data, which provides insights that can be fed into the system as it again refines itself.
The customer is the focus of the process, Sridhar explained. The entire purpose is to improve the experience for the client, and the client’s clients, through digitization. That means intelligent systems that can be made smarter over time.
There are several related benefits to this experience. The platform becomes an integrated development tool upon which more customer-centric applications can be built. Other customers benefit too, because the needs of one bank may be similar to another one.
With a solution already developed, Vuram is much closer to customizing the answer for that second customer, which shortens the time to market.
“It is basically an account onboarding kind of solution where any bank or financial institution can use it, which would have all the basic processes built into it,” Sridhar explained.
“Now there will be a difference in how RBC does it versus Scotia Bank does it versus TD Bank. But then the customization will not take you as much time as it will take you to build the end-to-end solution…I put in, say, two to four weeks of customization effort, and I’m ready to go live as opposed to building a solution that would probably take me about 16 to 20 weeks or probably even more.”
Human touch
While we have been forced to increase digitization over the past two years, Sridhar cautions many people now more than possibly ever need to interact with other people more than yet another system.
Vuram prioritizes making interactions as authentic as possible through AI-driven chatbots and natural language processing, which helps with sentiment analysis.
A goal is to automate the less productive and more repetitive work done by humans, leaving them the more highly skilled and motivating work, Sridhar said.
For example, bots can scan a passport to fill out application data, saving humans time. Individual customers are building bots for many reasons, and they are repeatedly improved through AI.
It also allows less technically proficient staff such as those on the business side to build applications for their sectors much more easily.
Sridhar said that most of the automation comes from robotic process automation (RPA) and business process management (BPM) tools.
RPA tools for legacy systems
RPA tools help when companies struggle to link their legacy systems like mainframes. AI then comes in to drive intelligent automation.
As natural language processing does with sentiment analysis, machine learning allows systems to be taught what a human has done. Say there are 10 types of high-risk applicants a bank typically rejects. The system can be trained to identify those, and as exceptions arise, integrate them to make itself more accurate.
Large companies are seeing the value in these processes and are rushing to build their offerings, Sridhar said. Google has DocAI, while Abbyy and Tesseract are other options. The opportunity is massive but must be considered together with how fast a business can adapt.
“Some clients work in regulation-driven industries where the regulations change quite frequently, and they want to be in a comfort zone where they are comfortable and confident that their systems work really well,” Sridhar said.
Tech linking tech
Another consideration companies had to worry about more in the past is how new technologies complement their existing stack. Now, much of that linking is done by the technologies themselves. And with more companies using technologies from RPA and BPM providers, what works for those providers will also work for the clients.
“In my opinion, it’s a good direction to move because all the products want to be self-sustainable, or they want to be kind of a thing where we just take it, and everything is there,” Sridhar said.
“It has all the capabilities, and all I need to do is just plug into my systems, and then it will start working fine. And that is where we are moving towards.
Silos are still a challenge, but primarily for larger companies. The smart ones of all sizes are already serious about hyper-automation. They want to build bridges between departments, especially as they seek to leverage the oceans of data they possess.
“Bringing structure to this data is one of the advantages that you can get by going into this hyper-automation, and people who realize that are the leaders,” Sridhar said. “They are the ones who are making a move, and they’re well ahead of the curve in terms of the competitors.
“It helps you in departments where you didn’t realize that it would help you.”
