Adopt Artificial Intelligence to improve operational efficiency in financial services sector

Vishwesh Padmanabhan

The explosion of emerging technologies such as artificial intelligence (AI) is dramatically changing the way businesses operate today. As businesses collect more and more data, the need for solutions to drive true value from that data grows in importance. AI, in conjunction with big data and analytics, can deliver that baseline value and go beyond traditional solutions to find deeper insights.

In India, banks are fast moving in this direction and deploying AI-powered chatbots for their operations to gain better insights into their customers’ usage patterns, offer customised products, help in detecting fraudulent transactions and improving operational efficiency amongst others. There is no denying that AI helps banks nurture their relationships through better interactions with their customers however, not without challenges.

The financial services sector has seen a vertical rise in adoption of AI technologies. As per KPMG estimates, the global AI spend is growing at a CAGR of 46 per cent (2016-21) with banking accounting for 18 per cent of the total spend.

While global financial institutions are focusing on use cases from front to back office, in India that scale is yet to be achieved. The push from government to conver public data systems and movement of banks towards open frameworks is extremely encouraging. Moving on across sub-segments within the financial services sector in India, the applicability of AI has primarily been seen in the in the front office. Several use cases in sales & distribution, customer front-end operations like chat bots and product customisations, have been implemented by leading banks and other financial institutions. That said, AI has seen fragmented efforts in back office areas such as fraud detection, fraud and AML (Anti Money Laundering) investigations, anomaly prediction and portfolio investigation.

From a financial services compliance standpoint given the dependence of AI driven applications on data, regulations around collection, storage and processing of data are critical to comply. For instance, RBI’s (Reserve Bank of India’s) data storage localisation regulation mandates storage of data by multi-national payment companies pertaining to Indian users within the country. Hence, for instance, a fintech payment application intending to use data of Indian users to offer AI-based transaction recommendations, needs to comply with such data storage rules in India. In the future, data privacy laws and regulations similar to GDPR, might pose stringent compliance requirements for companies processing large amounts of data and sharing data across multiple geographies.

In addition to increased compliance, plugging AI into financial services applications comes with its own set of challenges for solution providers. From building credit history to customer profiles, institutions need to build rich, organised and meaningful customer data, for AI-driven solutions to be deployed on a large scale.

Moreover, AI systems are intrinsically designed for a particular language, and continuously ‘learn’ and improve in that language. Given the linguistic diversity of the customer base in India, building a solution catering to a wide customer base, poses an additional challenge. Lack of skilled developers and data scientists in the Indian market is another big challenge in sustained development of AI-based applications in the financial services sector.

In order to build robust and meaningful AI offerings, financial services institutions must plan a well-designed and secured system of collecting, organising and archiving the right data. There is a need to collaborate with ecosystem partners that will help complement existing offerings and in return, provide a validation platform for nascent solutions. This can prove to be a low investment, high-return initiative, for a large financial institution that might not be agile enough to build and deliver such a solution.

Even though AI led capabilities are well-recognized by banks and financial institutions, they must move beyond fragmented efforts towards a sustained program to inherit AI in their products, services and operations. Use case driven proof of concept is a great way to start this journey with data-driven outcome.

The author is Partner, CIO Advisory, KPMG in India.