Kamal Misra, Senior Director and Head of Banking, Capgemini Invent India
Humans needed an alibi for efficiency as bots came calling and laid claim to the psychological frontier called “conversations”. Through an eclectic mixture of structure, design and innovation, these conversational agents or simply put, intelligent bots have taken the world by storm. In common parlance, the term intelligent indicates that these bots are engineered with speech recognition, natural language processing and machine learning capabilities to establish the quasi-human stereotype. While conventional bots chug along a mesh of mapped out conversations steered by a decision tree framework, the more advanced versions typically dip into a legion of Generative Artificial Intelligence (AI) and Natural Language Processing (NLP) conduits to synthesize responses.
These so-called Conversational User Interfaces (UI) and Intelligent bots have brought about a significant transformation in the banking industry. The emergence of these technologies has enabled banks to automate their processes, enhance customer service and experience, and reduce costs. Intelligent bots are gaining widespread popularity in the banking industry as they are designed to cater to a broad range of customer needs while being user-friendly.
A leading research firm got to unravel that the average cost for a phone steered transaction is around $2.50, while the same done online could be pegged at as low as $0.17. Bots peppered with low development costs and high data ingestion, processing and synthesizing capabilities have a considerable advantage over other digital media when it comes to efficient customer engagements.
Multilateral innovations in this space have revolutionized customer service in the banking industry. Intelligent bots can provide customers with instant and accurate answers to their inquiries without the need for human intervention. They can also assist customers with a wide range of services from checking account balances, transferring funds between accounts, and paying bills, to providing personalized recommendations such as ways to save money on fees and interest rates.
The advent of sprightly social media based conversational interfaces into monolithic banking structures has not only spawned “super apps” espousing the cause for well-rounded customer propositions, but also infused energy into an always-on, hyper-personalized advisor-inside-pocket wizardry. More power to robo-advisors and personal financial managers, as these bots fight off disintermediation, in that, they allow a 24×7 ever-engaging platform for banks to push their proprietary offers and recapture mindshare.
A case in point is Erica, Bank of America’s intelligent bot that had taken a herculean effort to be given shape and tested, has gone on to graduate with flying colours. Erica using advanced machine learning (ML) and cognitive messaging can adapt to complex and variable conversations with the customers. The chatbot can go as far as talking about cashflows, bills, transactions log, balance tracking, offers recommendations with discernible ease, while relaying back to the development team challenging issues such as navigation and unanticipated events. The bot is designed to improve customer experience, enabling them to access services faster and in a more personalized way.
Banks can use these technologies to automate tasks such as account opening, loan processing, and fraud detection. Also in the fray are tasks pertaining to lead generation, application generation and processing, bots can help banks reduce costs and enhance efficiency while freeing up staff to focus on more complex tasks. Bots backed by the ever-evolving AI / NLP capabilities, engage with customers right from the pre-approval phase to documentation and onboarding with necessary conversations throughout the stages, thus allaying any risk of abandonment in the mind of the prospective customers, establishing the first stamp of loyalty.
Capital One’s Eno goes a step ahead. Built entirely on the cloud, Eno’s Natural Language Processing (NLP) system with the help of complex deep learning algorithms, has been programmed to decipher when customers enquire about account balance in umpteenth number of ways. A research pilot by Capital One discovered 2200+ diverse ways or expressions that a customer can intend to use to ask for an account balance, using different phrases, abbreviations, spellings and connotations. Eno’s NLP training included the ability to recognize and imbibe new misspellings and abbreviations that it previously did not get trained on. It also assists with credit card activation, fraud detection, account management, and offers recommendations on how customers can improve their financial status. Eno enables customers to access services faster and eliminates the need to visit a bank branch, thereby improving customer experience and reducing costs for the bank.
COIN (short for Contract Intelligence) from the stable of JP Morgan Chase investigates 12,000+ contracts annually, saving the bank an estimated 360,000 person hours.
Caixabank’s chatbot with its advanced AI engine claims to understand and respond to 1500 questions in different language. As per the bank, the chatbot received on an average 50,000 questions per day through 2022 and has purportedly given an optimal response to 85% of the total questions posed, leading to no further human intervention.
Furthermore, intelligent bots are transforming the way that banks provide personalized recommendations to customers. These bots use machine learning algorithms to analyze customer data such as transaction history, spending patterns, and credit ratings, to provide personalized recommendations that are tailored to each customer’s unique needs.
Wells Fargo’s Control Tower is an intelligent bot that offers services such as personalized financial advice and recommendations on how to save money on fees and interest. Control Tower can also help customers to set up savings goals and track their progress over time, further enhancing customer experience.
Morgan Stanley plans to roll out a low-cost digital only avatar of it’s next-best-action system aimed at providing pre-emptive investment options to customers through its network of 16,000 financial advisors. This class of augmented relationship management advisory helps customers receive personalized alerts tied to specific life events, such as a child’s birthday or illness.
In addition to improving customer service, automating routine tasks, and providing personalized recommendations, intelligent bots are transforming the way banks engage with customers. These technologies can be used to provide customers with personalized experiences, such as targeted marketing campaigns and customized financial advice.
Ally Bank’s Ally Assist is an intelligent bot that provides customers with personalized financial advice and recommendations. Ally Assist uses machine learning algorithms to analyze customer data and provide personalized recommendations on how customers can improve their financial status. Ally Bank can also use Ally Assist to provide targeted marketing campaigns to customers based on their spending patterns and financial goals, thus enhancing customer experience and driving customer loyalty.
However, despite the many benefits of intelligent bots in the banking industry, there are also potential risks and challenges. One of the key risks is the potential for fraud and security breaches. Banks must ensure that their conversational UI and intelligent bots are secure and protected against hacking and other forms of cybercrime. This is crucial as these technologies often hold sensitive customer information that could be compromised if not adequately protected.
Another potential challenge is the need to ensure that these technologies are accessible and user-friendly for all customers, including those with disabilities and those who are not familiar with new technologies. Banks must provide appropriate training and support to ensure that all customers can use these technologies effectively, which will help reduce customer frustrations and increase adoption rates.
