The transformation of RBL Bank into a digital hybrid bank
Using an AI-powered model, the bank loads 20-30 per cent less money in the ATMs, without changing the customer demographics
“Banks are witnessing a substantial tranformation and digital banking wave, thus it is compelling to join a bank at this stage,” says Sankarson Banerjee, Chief Information Officer, RBL Bank. Banerjee had joined RBL Bank in October 2018, and is accountable for spearheading the IT functions, strategies and designing the digital roadmap for the bank.
Banking has evolved into a 24×7 business, and banks are orienting their people, processes and technologies to build powerful core banking applications with personalised offerings. “System reliability and internal systems have to be 24×7 available, recoverable, resilient which require a significant changes in our infrastructure and designs,” explains Banerjee, adding that banks are the enablers now, rather than the controllers of technology, forcing enterprises to rethink their technology stack, distribution strategy and channels.
Recent technology developments
At RBL Bank, Banerjee is spearheading initiatives to transform the bank into a digital hybrid bank, which is a composite of traditional infrastructure and digital frontend. It will enable them to significantly improve customer delivery capabilities.
They are heavily investing in cloud-friendly technologies and for front-end, they are investing in microservices and micro-segmentation. These investments offer a lot of flexibility, without increasing the cost of deployment and change.
“Since the bank wants to build standard servers and services, we are focusing on the cloud first or cloud-only technologies. Then, we are totally working on the new programming languages, i.e, Java scripts, Kotlin, etc., which are lighter and faster than the traditional languages and have turnarounds of 5X and 10X more than the traditional languages,” states Banerjee, describing his IT roadmap.
Automation for everything
Further, Banerjee speaks about the key drivers leading to an increase in automation and AI adoption. Banks with multiple manual processes and interventions may not use necessarily robotics but are using process automation for straight-through processing.
The cost of service is the key problem in banks. “With RPA, banks can implement the complex and sophisticated loans and dimish the human intervention. This kind of automation does not require higher reliance on hiring skillsets, as the skilled employees are harder to hire and pricey. Whereas Robotics Process Automation is depreciating the cost, initially, the cost is high but gradually the cost reduces,” he shares.
Banerjee is also leveraging AI to make the customer experience seamless. He is closely studying customers and identifying customer trends and behaviours, driven by large scale data analytics. AI is used to identify patterns and predict customer demands based on their choices.
Personalised insights
RBL bank is using an AI backboned model to identify the optimal loading of ATMs. The AI solution analyses the customer’s withdrawal patterns from ATMs, in order to timely refill the machines. “The ATM should never be empty but at the same time, money should not be lying wasted in the ATMs, which is an expense for banks and a tricky balance to achieve. The AI-powered model has proved quite successful, we are able to load 20-30 per cent less money in the ATMs, without changing the customer demographics,” Banerjee explains.
It tracks down the different sources of data and applies AI to discover the partners.
“Hyper-personalisation has a lot to do with AI and analytics, it’s difficult for an individual company to lay down the hyper-personalisation model standalone, hence, we collaborate with academic institutes. They have the brain power, and students who think laterally and once the theoretical model has started to form its shape using the data, we start drilling it out to apply it on our businesses,” he adds.
The cost to the server must be low, as banks operate at a very narrow cost to server ratios, hence the cost of delivering these servers must be as efficient as possible. “We are using automation, digitisation, and hyper-convergence to reduce the cost,” Banerjee concludes.
