Parry logs in to his digital banking app where he is authenticated using his front facing smartphone camera for adequate security to transact safely. The app suggests different preapproved financial products which he is eligible for based on his profile, and at attractive pricing. He voices out his queries and gets immediate and accurate response from the app thereby empowering him to make informed decisions. Further, the app can even perform non-financial and financial transactions in certain cases where the customer has given authorization to do so. The machine learning (subset of AI) algorithm finds patterns from the data and determines if a transaction is permitted or not, and instantly process or rejects the transaction. In certain cases, the algorithm processes the transaction under assessment and flags any inconsistencies or anomalies. Basis feedback from humans on the anomalies identified, the program readjusts its logic dynamically.
In the use case above the app or a smart virtual assistant (SVA) uses visual recognition (VR) technology to recognize the customer. Similarly, it uses speech recognition technology to capture and interpret information thereby emulating the sensing aspect of human behavior. It uses advanced analytics powered by big data, cloud computing and machine learning to offer personalized products to the customer thereby emulating the thinking aspect of human behavior. It responds to the customer using natural language generation (NLG) technology providing insights and advice thereby emulating the action aspect of human behavior.
Smart virtual assistants are AI solutions that can interact, receive and deliver information and act on human commands. They are created using a combination of technology building blocks for AI powered solutions like machine learning, visual recognition, natural language processing (NLP)/speech recognition technology and natural language generation. Visual recognition technology transforms identity management. Speech recognition technology captures and interprets information and converts spoken language into machine readable format. Data, analytics, cloud computing and open source AI algorithms are the foundational building blocks for AI. Prescriptive analytics makes use of structured and unstructured data based on customer’s activities and interactions from various sources to make a personalized offering. Banks can collaborate with other participants in the ecosystem with automated business-to-business interactions to exchange data in a scaled manner using APIs and offer a full suite of products to customers through the open banking model. These technologies enable anytime, anywhere banking, offer 24/7 customer service and increased transparency of transactions thereby providing revenue generation and cost saving opportunities for the bank.
Banks can choose between platforms, applications and cloud services for deploying AI. It is generally accepted that a combination of all three is both practical and desirable. AI adoption accelerates banks’ digital transformation agenda of becoming more agile and customer centric. It frees bank staff for creative thinking, complex problem-solving and helps them focus on business strategy. Interaction with smart virtual assistant empowers the relationship manager as well by providing the details of the interaction and helping her / him find new business opportunities. Basis feedback from the relationship manager, the program adjusts its logic dynamically and acts on its own thereafter, recollecting previous interactions that have a bearing on the current decision to make relevant offers to customers in real-time.
Thus AI improves the overall experience for both employees and customers of the bank and any business in general by way of fewer errors, higher efficiency and better decision making.