Private sector banks looking to leverage robotics, AI
- Air quality in Delhi today best in over a month: CPCB
- Zimbabweans celebrate Robert Mugabe’s imminent exit at mass rallies
- India’s Manushi Chhillar crowned Miss World 2017
- Farmers to participate in two-day protest in Delhi from Monday
- Gujarat elections: BJP issues second list, names candidates for Congress held seats
Mumbai: Leading private sector lenders such as HDFC Bank and ICICI Bank are experimenting with robotics and artificial intelligence to improve efficiency and stand out in a market where banks are becoming more and more standardised in terms of technology and services offered to customers.
While HDFC Bank launched its humanoid Ira in January to assist customers at a Mumbai branch, ICICI Bank said in September 2016 that it was using robotics to reduce repetitive work for employees, allowing them to be delegated to duties requiring human interaction and involvement.
Developed by Kochi-based Asimov Robotics, the humanoid Ira is currently deployed at HDFC Bank’s Kamala Mills branch in Mumbai. In the next two years, the bank aims to introduce about 15-20 such humanoids across branches in India. The humanoids will primarily be stationed at welcome desks and will guide customers to relevant counters.
But the application of robotics in banking is not limited to this, said Nitin Chugh, country head of digital banking at HDFC Bank. The private sector lender has also introduced a chatbot, Eva, on its website, which helps customers get instantaneous responses on products and services.
Eva is an electronic virtual assistant or automated programme designed to help users communicate with businesses.
Apart from this, HDFC Bank is also trying to use robotics in certain procedures in its back-end to ensure a quicker turnaround time and error-free execution. “Our business follows a lot of processes and is highly regulated. Because it is that way, the processes become predictive and repetitive. Now when that happens, there are elements in that process which can be ‘robotized’,” Chugh said.
Once a process is “robotized”, the bank can then go ahead and introduce some elements of machine learning which can help the programme optimize itself and become better, thereby increasing its use case.
Not every such application is successful. HDFC Bank tried to introduce a bot to categorize customer emails and allocate them to relevant departments. But the accuracy rate on this was very low, which made the whole process pointless, Chugh said.
ICICI Bank, however, has been successful in its email bot experiment, according to Anita Pai, senior general manager and head, operations. ICICI’s bot, used to sort out customer and distributor emails on the status of transactions and similar things, has helped the bank cut its response time, Pai said.
“This process is now completely free of any human engagement, as the bot is able to pick up which department is responsible for said transactions and find out the status,” Pai said.
Turnaround time for processes such as disputed transactions at automated teller machines (ATMs) has also been reduced considerably, Pai said. It used to take the bank about five to six hours to fully address a customer’s dispute request. This time has been reduced to about one hour with the help of a bot that can handle multiple processes at the same time as opposed to an employee who performs functions sequentially.
When ICICI Bank started introducing robotic software to deal with back-end procedures, it transferred about 200 processes to the software. One process was collecting and sorting 15G and 15H forms which help customers save tax on their returns from fixed deposits. As customers could submit physical forms, the process became a big task for employees in the months of April, May and June.
“Once the employees were done with their day’s work, they would be handed a stack of these forms, which needed to be sorted and verified against the PAN number. It was a very taxing job, leading to late working hours,” Pai said.
Once the process was moved to the robotic software, the bank’s employees were no longer required to stay late to complete it, Pai said.
At present, more than 500 of ICICI Bank’s back-end processes utilise robotic software, freeing up a number of employees for duties requiring human intervention. The bank expects about 1,000 such processes to be shifted to robotic software over the next 12 months, leading to about 40% of its current processes becoming fully automated.
ICICI Bank has also introduced a chatbot on its remittance portal Money2India, which has helped reduce customer queries on phone and email by a third.
Introduction of robotic software has also provided flexibility to ICICI Bank to switch on or off an entire process. In the case of the 15G and 15H forms, the income tax department has recently allowed the use of electronic form submission, which makes the bot defunct.
“But in the process of using these softwares we have also learned a thing or two about some other process which are now being fixed. This means that the bot has paid for itself,” Pai said.
However, automation has not led to any shrinkage in the workforce of the banks. Both HDFC Bank and ICICI Bank have continued to hire new people and have no intention of cutting any jobs, Pai and Chugh confirmed.
“We are clear that wherever there is scope to improve interaction with the customer is where we will use robotics and automation. We are trying to only augment the way our employees offer services, not replace them with machines,” said Chugh.
According to Vivek Belgavi, partner, FinTech leader at PwC India, while reducing cost has been a major reason for adoption of robotic processes, improving experience is also a big push.
“Going ahead, we may see rationalisation of headcount in the more monotonous work areas, banks may hire more people in the customer facing processes,” Belgavi said.