How AI is Disrupting the Banking Sector in India

Nitin Chugh, Country Head - Digital Banking, HDFC Bank, and Sachin Jaiswal, CEO, Niki.aiArtificial Intelligence, for beginners, is the ability of machines to think on their own and execute a task without human intervention. It is achieved through technologies like Machine Learning, NLP, Deep Learning, Speech Recognition, Image Analysis, and others. by collecting, cleansing and analysing massive amounts of data.

Keeping aside the Sci-Fi movies, you realise that there is more to AI than just humanoid robots and machines ending up taking over humanity. The field is still being explored and worked upon for it to reach its true potential. However, much like Industrial Revolution automated human labour to a great extent, thus increasing production and reducing costs, AI is being used to automate digital processes in a number of sectors including but not limited Health Care, Marketing, E-commerce and Finance.

Banking is one of the sectors which is seeing rapid adoption of AI, as banks are being proactive about improving customer experience by leveraging the abundance of data which the business revolves around. Be it lending, insurance or wealth management, AI technologies have shown potential of making them all better.

HDFC Bank OnChat, which is created in partnership with, is one of the notable examples on success of AI in banking. The Facebook Messenger bot lets one transact for utility bill payments, entertainment and travel services like movies, cab and bus booking. The bot has seen a 160 percent m-o-m growth in the number of transactions, proving the chat experience to be convenient and easily adoptable for the users. The total number of messages exchanges by the bot has crossed 2.4 million. Number of repeats, which is close to 34 percent, further bolsters the confidence in conversational banking. The Messenger chatbot also opened up yet another channel for customer acquisition, the current percentage of non-HDFC bank customers using the bot stands at around 25 percent.

From a customer standpoint, starting with a millennial who finds finance boring, to those in their mid 40s who find wealth management stressful, AI comes to the rescue by making the experience naturally pleasant and taking away the troubles of number crunching.
Why Banking, Specifically?
Banking is data-intensive, and analysing the abundant available data, meaningful patterns can be drawn using AI. With so much of advancements in technology, today's consumer has got a taste of ease and convenience. He/she doesn’t want to stand in a queue or wait for hours to connect with an agent and get an answer for a query, to go through number of steps and paperwork to simply buy an insurance, or to get analysed for creditworthiness over weeks of time.

AI will take away all the monotonicity there is when it comes to banking, and use the huge amount of structured as well as unstructured data involved in banking processes which is being generated with exponentially increased mobile and internet usage, to improve the customer experience and increase efficiencies.

Here’s how banking can be redefined using the power of AI –
Lending: Not going far, only over two percent of Indians have access to credit. And access to capital is required to foster growth in an economy. It has been a cumbersome task for banks to analyse an individual for creditworthiness, due to the lack of credit history.

Customer Service becomes more important in banking sector due to involvement of hard-earned money and lack of education among customers

However, new age alternative lending startups are leveraging Big Data, Data Science and ML to analyse spending patterns, direct, indirect, publicly available and behavioural data of a customer over 10,000+ data points, to have an insight on his/her creditworthiness. This also eliminates the need of a lot of paperwork, and allows those out of the financial fold (self-employed and students) to obtain credit.

Insurance: Insurance is a data heavy sector as well. It is not feasible to manually analyse TBs of data and draw patterns which might help in fraud detection or claim management. Deep learning techniques for image analysis can help in automated repair cost analysis in damaged vehicles, with the help of predictive models. Hence, the turnaround time which plays a crucial part in customer experience, can be reduced. On the other hand, graph analysis on publicly available data (social networks) can help draw fraud detection patterns.

Customer Service is always at the forefront for any business, and it becomes all the way more important in banking sector due to involvement of hard-earned money and lack of education among customers. Conversational agents such as chat bots or voice bots which utilise Natural Language Processing to read, process and understand text/speech, have proven to be a success for many banks where most of the queries are being automatically answered by a bot, reducing the human capital required, and increasing customer satisfaction due to consistent and quick service.

Sales can be increased and the experience made better by AI-based bots which take the customer in the journey from discovery, to choosing to purchasing in the most natural way possible — conversations, which involve a personal touch even when a human is not present on both the ends.

Wealth Management: Machine learning based predictive models which constantly analyse data and draw patterns can help potential investors choose the right product for their portfolio, and give insights on price fluctuations in the future. The AI-based systems can analyse the amount of salary (even if it varies) you receive every month, the spends, and savings to predict future behaviour and tell you that you’ll need to save Rs.50 more per day to be able to buy the bike.

Although these are just some examples on how banks can leverage AI to improve overall efficiency of the organisation and to provide great convenience and value add to their customers, it will still be too soon to judge the sheer impact the technology can have on the lives of millions. Rapid adoption, which leads to rapid innovations, is definitely something that will be interesting to see in the coming years.