Banks were already marking digital transformation as a top priority as they faced competition from challenger banks and fintech startups who offered faster and more convenient options to their customers. But the pandemic, the rush caused by financial uncertainty and the increased need for online services has pushed that need to a higher level.
Traditionally, banks have been slow in adjusting to new technologies. While having a long history of tech infrastructure going back to mainframe-based systems, that same established foundation is now acting as a barrier, since managers are unwilling to abandon tried and tested systems for untested advancements. And the topic here is financials, involving millions of dollars and making the risk of failure extra high.
There is also a mindset problem, as senior executives with 30-40 years of experience have their own views of how the industry should work. Some private banks even lack an online system to begin with, as their clients typically are wealthy and old people who are not fond of technology.
But this is all changing, as younger millionaires (sometimes inheriting the fortunes of their parents) are stepping in, and younger executives are being appointed to senior jobs. Add a pandemic which forced banks to process two decades of small-business loan applications in a single month, and we can see that change has to come.
Technology can help banks in a number of ways: from using AI for fraud detection and risk management to international transactions with the aid of blockchain technology. But one of the most profound impacts is using conversational AI for marketing and customer support.
Conversational AI is a technology that banks have already tried, but with varying degrees of success. The benefits are obvious: this is how banks interface with customers to offer them personalized services, assist them with loans or ensure they are happy clients and will stay longer with the bank. AI can be a huge assistance, as it can engage directly with customers and respond to their queries, making the job much easier for the bank’s operators. AI may also engage with the operators and give them hints on what to respond to the calling customers, or provide them information in real-time, helping to resolve the case faster and keeping the clients less on the line.
Earlier conversational technologies such as chatbots were promising but failed to deliver the expectations: they were only made for simple scripts, and a slightly different conversation would render the bot incapable of offering meaningful answers. Natural Language Processing-based AI technologies are much better at the task as they can follow the conversation with the client, remember the context of the call, detect user intents, and resume a topic discussed earlier.
However, barely using NLP is also not enough, as human conversation is simply too complex most of the time. At Kore.ai, we are merging three technologies to achieve state-of-the-art conversational AI, including Machine Learning, Fundamental Meaning, and ontology-based Knowledge Graphs. This approach enables us to combine the strengths of all and overcome the weaknesses of each of them to deploy a model that is very capable of engaging in human conversation.
Some of the banks that have tried Kore.ai report that over 50% of their calls can now be taken care of completely by the AI. At times when customers are rushing to call centers, having an AI system in place that can take parts of the loads can mark the difference between a system being overwhelmed, and one that can keep responding to the customers.
Digital transformation is inevitable, as it is being required more and more by customers—not just because of the pandemic, but also because of changing demographics. Those who adopt it sooner will attract more clients.