As banks are seeking to embrace AI based applications to enhance customer experience, improve compliance and drive cost efficiencies, AI is now becoming increasingly naturalised

Artificial Intelligence is not a new phenomenon. It has been in vogue for a while, although generally associated more with complex algorithms that were related to high risk and high-value transactions, as the underlying cost of technology was expensive, and the availability of necessary data was scarce. Both of these entry barriers have now become irrelevant, particularly with the advent of the Cloud and Big Data.

Industry estimates peg digital data that is globally generated every second day now as more than what we had produced collectively in the whole of a year, 20 years back! And this is not without reason. Even as banks are increasingly naturalising AI with permanent citizenship as part of the technology landscape, the appetite for gallons of digital data and their consumption will only grow from here.

AI is already in the house!

We are already experiencing the advent of AI in almost every walk of a bank, many of them camouflaged to be just a simple technology advancement. Let’s look at 10 simple AI-enabled applications that we are experiencing already, and which are now integral to most banks today:

  • Customer engagement: Virtual assistants at contact centres and enabled chat-bots help free agents from high frequency and high-volume enquiries that are relatively mundane, regular and low on complexity. These enquiries are now driven by AI. One could experience them both in the regular customer service functions and across specialised business units such as wealth management and trade finance operations.
  • Conversational bots: One of the greatest benefits of Machine Learning is evidently observed in speech recognition, a science that significantly benefits the quality of customer engagement. The plethora of information across banking systems, call centre patterns, and social media helps the machines to continuously learn and improve the quality of their responses.
  • Smart Servicing: Banks find it ‘cool’ to deploy humanoid robots to serve their customers, providing an air of ‘science fiction’ like experiences at branches. While facial recognition helps meet-and-greet customers at branches, AI enablement can also allow for better experiences on other channels – such as frictionless payments on mobile smartphones.

We are already experiencing the advent of AI in almost every walk of a bank, many of them camouflaged to be just a simple technology advancement

  • Personalised communication: A fair degree of communication – including customer on-boarding, potential new offers, advice and campaigns related to services – is now personalised, even while remaining automated. This is on the back of AI engines that process the underlying data that make the bank able to be more precise and targeted – improving the relevance of communication and return on investment.
  • Personalised products: Be it in investment or trading products, or offerings from the bank for a loan or for a credit card, the benefit of reviewing the 360-degree engagement of the customer improves the quality of the appeal manifold. For instance, payments and spend analytics help in promoting personalised coupons or rewards programmes, which can be both time and/or geography-specific. AI-powered virtual advisers are also made available to offer ‘personalised’ advice to the customer. AI enables banks to ‘anticipate’ customer needs and align potential offerings around established preference patterns.
  • Credit Scoring & KYC: Getting the customer-validated for KYC requirements and building credit scores based on past experiences as well as multiple data sources, is another interesting application driven by AI engines, which is now becoming mainstream. AI-enabled platform help instantly picks up images from document scan and assesses the application based on credit scoring models with instant approval decisions. The models are designed to go through an auto-learning process thereby improving the quality of decisions based on the feedback loop, over a period of time.
  • Collections & Recovery: Well-managed collections and recovery models are built on AI-enabled early warning models and predictive analytics that help identify customers with a higher propensity to default. Successful models are then designed toward pre-empting delinquencies or defaults, with customised solutions that may be specific to each such customer.
  • Fraud detection: Real-time detection of fraud not only requires a continuous scan of the customer and transaction data but also leveraging other data points related to employees, and third party sources and correlating with patterns. The need for near-real-time shadowing of transactions and analysis in split seconds makes it difficult to execute manually, and the need for AI here is imperative.
  • Compliance reporting: The cost of compliance has a strong quotient of labour involved in executing repetitive report generation and confirmation activities, as banks need to file an array of regulatory returns. This gets even more demanding, where new regulatory requirements need to be constantly tracked. This is yet another universal candidate for AI adoption, as observed with several leading banks wherein NLP [natural language processing] and deep learning help with compliance with reporting regulations.
  • Cost efficiency improvement: Automating and streamlining simple, mundane processes such as mortgage processing, reconciliations, authentication, and customer verification are examples of AI-enabled process efficiencies that banks are beginning to embrace as mainstream.

While many of the above are use-cases that banks have been embracing around the world, the real adoption of these is a function of how well the bank is positioned to imbibe them in a seamless manner. A successful AI bank is not the one that has just invested in experimentation and has got a few use-cases rolled out. On the contrary, it is one that has been able to demonstrate scale, on a sustained basis. AI is not a one-time project, but a journey that banks need to internalise as a mission, with a long-term orientation.

The 4 pillars of AI wisdom

Here are some simple questions across 4 pillars that a bank’s leadership may need to ask of itself. Responses in the affirmative are invariably going to be a prerequisite for driving scale and sustainability.

Strategy: Do we have the right AI strategy, including a framework that allows for prioritisation between competing, and equally compelling applications that demand resources? Are the AI strategic objectives aligned and dovetailed with the bank’s overall strategy such as revenue growth, cost reduction or improved compliance?

Process: Can the AI applications be embedded in the existing processes? Is the enterprise well positioned to review, improvise, and drive the benefits to scale with effective design-thinking capabilities? Is the enterprise ready for accepting and institutionalising the change that comes along with the re-imagined processes across channels? Do we have the right governance processes to monitor the data security and the agile processes to build, test and deploy AI models?
Technology: Are the necessary tech-stack and the supplier/partner ecosystem required for driving the AI journey well in place? Do we have the necessary API connectors and ecosystem to support reusability? Are the AI solutions well integrated with the overall application ecosystem? Do we have the necessary data elements, with a ‘single source of truth’ to drive the AI application effectively? Have we tapped into the cloud infrastructure to provide the scale and resilience to sustain?
People: Does the bank have the specialist resources required for sustaining the AI journey? Do we have the willingness to experiment and have the culture that is required to drive this change, with the performance measurement to track progress in place? Are the team members trained to deal with agile ways of working and collaborative working models? Is the learning environment configured to provide the skills that need to be imparted on a sustained basis?

At the end of the day, artificial intelligence alone does not help, unless it is aligned with the overall strategy of the bank, and implemented effectively, duly reflected in the process efficacy enabled by contemporary technology and a performance-driven culture. The proof of the pudding, as they say, will ultimately be in the eating!

To read more such insights from our leaders, subscribe to Cedar FinTech Monthly View

Talk to our Consulting leaders about how we can add value
Contact us to make strategy & innovation work for you

Relevant CedarViews