Kalpesh Mehta, partner, Deloitte India, speaks about the transition AI is bring to the financial services industry:
Kalpesh Mehta
Artificial intelligence is no longer artificial, but in the realm of reality and an integral element of our daily lives, says Kalpesh Mehta, Partner, Deloitte India. “A blend of machine learning, cognitive computing and natural language processing, AI has a huge relevance in banking and financial services and the applications can range from improving workforce productivity, enhancing the customer experience, driving innovation in financial products, launch of new business models and ensuring regulatory compliance,” says he.
He explains that the range of applications of AI in banking range from anti-money laundering pattern detection, chat bots in managing customer relationships, algorithm-based trading for high frequency trades, fraud detection and driving customer recommendations based on customer preferences and past financial history. And Indian banks have started using AI – for example the use of chat bots for advanced customer services or use of natural language processing to help customers with financial and non-financial transactions or get in touch with the bank executives for products. The other uses are streamlining systematic and data focused banking operations for example in risk management and managing delinquencies, he adds.
CREATING NEW MODELS
Today, says Mehta, digitization of payments and using AI with payments data is leading to creation of new forms of transaction-based, close-looped lending models. And this can lead to penetration of credit to the MSMEs and other under banked segments with little or no access to credit. “The use of wearables for health monitoring is leading to new forms of insurance products where the customer’s behavior to maintain good health is rewarded with lowering of health insurance premiums. In the case of personal finance management, use of AI and robo advisors is leading to a better management of an individual’s personal finances and allows financial institutions to recommend financial products based on an individual’s financial goals and behavior,” he elaborates.
Mehta also believes that AI is definitely beyond chatbots and conventional interfaces. “It’s true that investing in the right AI technology can have a major impact on a bank’s operational efficiency, and that its success boils down to the customer impact above all else, and like any technological innovation, the best results will be realized only if they are improving the end user’s experience. That is a bare minimum, but AI will go beyond and lead to a bank or a financial institution creating new financial products,” he says.
DECISION MAKING POSSIBLE
He cites the instance of banks exploring the use of digital payments data or customer’s GST transaction data to make lending decisions. This will lead to a better penetration of credit in the country especially into micro and small enterprises. “The mainstream uses of AI. Hence, will be in areas of new product development, pricing of financial and insurance products, workforce productivity, fraud management, credit risk management, customer services and managing delinquencies,” he adds.
According to him, banks in India are adopting AI in a variety of forms – chatbots for customer service, innovation factories for development of AI technology and products that improve customer engagement, operations automation, hackathons for innovative product ideas and solutions through use of AI etc. “The future as we see will be that AI will go mainstream across the banking operations and in areas such as new product development, pricing of financial and insurance products, workforce productivity, fraud management, credit risk management, customer services and managing delinquencies,” he predicts.
Mehta maintains that use of AI can definitely lead to ‘cognification’, when AI is used to make specific decisions – for example in areas like micro-lending where the reliance on non-traditional sources of data such as payment transactions, tax transactions is higher. “In such cases, due to the high volume of transaction data to be analyzed, cognification leads to a system analyzing and making a credit decision over a human. With growth in digital payments volumes and GST related tax transaction volumes, cognification is the way forward in the lending process to make credit decisions. It is my belief that going forward with increased cognification, AI will be used in banking and financial services to make specific decisions,” says he.
BROAD USAGE PATTERNS
He outlines the use of AI in the banking and financial services sector as:
With digitization of payments, there will be a transactional trail for every digital transaction. Money launderers use innovative and creative means to mask their actions making the illegal earned money look legal. AI helps detect patterns across transactions that are otherwise difficult to detect. Banks will shift to AI based systems that are intelligent in detecting underlying patterns across transactions and ensuring AML compliance.
Trading systems deploy sophisticated algorithms and models that require inputs from wide range of data sources (financial markets, company databases etc.). With an increased trading based on sophisticated algorithms, the trading decisions are automated, and system based.
With the surge in volumes of online transactions and a focus on increased convenience to customers, banks are using AI to increase the customer experience, an early improvement in fraud detection and keeping the false positives lower.
There can be 3 forms of customer engagement that can benefit from an increased use of AI. Firstly, in use of chatbots where the human chat can be replaced by a system yet being able to deliver greater personalization and customer relationship management. Secondly, in the use of virtual advisors, ie product recommendation engines that use the customers’ financial goals, financial behavior, financial data and personal data to recommend financial products. Thirdly, in taking financial decisions to provide a customer with an instant experience, for example, in lending decisions.
IRA 2.0, an advanced humanoid from HDFC Bank
AI used in risk management can shift the focus of risk managers to a proactive risk management role. AI based operational risk platforms evaluate the customer data from a wide variety of sources, identifying the complex patterns within the data leading to an accurate risk prediction. The risk models learn from every new data set and improve the prediction accuracy over time. A bank’s risk functions will see application of AI in areas such as credit cards, SME lending, underwriting decisions etc.
AI will help regulatory compliance teams of banks to manage key compliances such as KYC, AML, rogue person detection, trade monitoring, transaction monitoring etc. The fact that digitization of banking systems will lead to a surge in the volume of data and hence there will be an increased regulatory focus on the digital transactions, which will lead to an increased use of AI in managing regulatory and compliance requirements.
WORKER EFFECTIVENESS
He also foresees that process applications will incorporate AI into an organization’s workflow to either automate processes or improve them by augmenting worker effectiveness. “Automated voice response systems have been used for some years now to replace human customer service agents for first-tier customer support. The Hong Kong subway system employs AI to automate and optimize the planning of workers’ engineering activities, building on the learning of experts. Insight applications harness advanced analytical capabilities to uncover insights that can inform operational and strategic decisions across an organization,” he says.
Mehta outlines the extensive benefits of using AI as: “Changing customer expectations in an increasingly digitized world are disrupting the industry at large. For an industry rooted in traditional orthodoxies, new disruptive forces are threatening long-standing entry barriers. Those who survive and grow will reinvent themselves and take faster steps toward transformation. Organizations can realize costs savings through the effective use of AI. Other potential benefits – from improved flexibility to higher employee morale – can extend the value of cognitive automation.”
MAJOR ADVANTAGES
Some of the perceivable advantages, according to him are:
- Decreased cycle times and improved throughput: perform tasks faster making 24×7 operations possible
- Flexibility and scalability: once a process has been defined as a series of instructions it can be scheduled for a particular time and deployed to perform it.
- Improved accuracy: programmed to follow rules and do not make typos.
- Improved employee morale: The tasks and processes most suitable for automation are typically the most onerous and least enjoyed, if humans are relieved of these tasks they can focus on more important and rewarding work.
- Detailed data capture: The tasks performed can be monitored and recorded at every step, producing valuable data and an audit trail that can support further process improvement and help with regulatory compliance.
Finally, does he think AI will lead to replacement of humans in certain operations?
“I don’t see that happening. If anything, AI will augment the human role in banking operations. It will free up the human time spent on mundane or time-consuming tasks and release humans to focus on more productive operations like managing customer relationships which require a more emotional and personal touch.”
