Listen to this story
India has seen the banking sector adopt data analytics and AI faster than most countries. As per a survey by PwC-FICCI, AI applications are estimated to help banks make potential cost savings of $447 billion by 2023.
In an exclusive conversation with Analytics India Magazine, Sonali Kulkarni, lead, financial services, Accenture in India, shared her insights into the world of AI transforming the banking industry.
AIM: Banks have been using data and AI for some time now. What has the adoption in India been like so far, and where do you see this headed?
Kulkarni: Despite a slow start, most mature banks in India have started their data and AI adoption journey by making foundational investments in analytics use cases, data lakes and customer journey digitalisation. They are using data and AI to improve decision-making across the banking value chain. For example, machine learning is being used to improve cross-sell efficiency, targeted digital marketing, estimate core balances, improve capital efficiency, and for smarter underwriting decisions.
During the pandemic, there was a surge of investments in data and analytics-driven risk discovery and mitigation to get early warnings on market and credit risk, forecast liquidity needs, identify delinquency patterns, improve collections strategies, and also for fraud detection. AI is also being used to support the Know your Customer (KYC) processes, for generating credit appraisal memos and regulatory compliance tasks such as filing suspicious transaction reports.
We expect banks to continue to innovate on new cross-sell and profit optimisation models to tap into their rich data reserves as well as pivot to building enterprise-wide intelligent automation and AI platforms. The need to compete with digital native players, drive superior customer experiences and extract better value from digital investments will drive innovation in this space. Lastly, responsible and explainable AI practices will be a key focus for banks.
AIM: How can banks best approach AI-enabled business transformation so as to derive higher returns on investments?
Kulkarni: First, AI should be applied with a business intent rather than to fulfil a technology goal. Second, banks need to think of longer-term business outcomes and hence, must go beyond the siloed proof of concepts and apply AI across the organisation. Leadership commitment is vital to achieving this.
Banks need to build an AI core that integrates foundational capabilities across the organisation – such as cloud-native data lakes, AI services, data platforms and tools, robust security and governance. It is also key that banks do not just limit themselves to technology interventions but also focus on building data, digital and cloud skills in the workforce supplemented by new operating models, and data-driven ways of working.
AIM: How can banks use AI and analytics to unlock value in relatively untapped segments such as treasury operations and corporate banking?
Kulkarni: Treasury operations can benefit from curated AI algorithms that can analyse large volumes of data with greater accuracy, and thereby help treasuries manage risk and forecast liquidity better. They can also manage cash more effectively and efficiently, and implement controls in a timely manner.
AI and analytics can play a pivotal role in coverage management and improving client servicing in corporate banking. AI-driven insights, when applied to portfolio management on an ongoing basis, can enable proactive monitoring. They can also help the bank pre-empt customer queries and servicing issues and empower relationship managers to take proactive or preventive steps. AI driven nudges can also help relationship managers prioritise key tasks.
Lastly, the true value that data and analytics can drive is to offer the bank a 360-degree view of the customer and break the silos between the corporate banking and retail banking businesses. Data and analytics can address needs of a client across their entire ecosystem, including at an organisation level (corporate banking), for their suppliers (SME banking) and their employees (retail banking). This approach would unlock value for customers as well as for the entire bank.
AIM: New entrants to the financial services sector may not have significant customer data to derive insights from. How can they get started?
Kulkarni: Leveraging data is an ongoing journey, and every organisation is at a different stage of this journey. New entrants to financial services that do not have significant customer data can begin by focusing on low-hanging fruits, such as heuristic analytics, where insights from the available data are driven through business judgement or predictive insights from an expert opinion. And as they accumulate more data, sophisticated analytics can be leveraged to replace expert judgement with machine learning models. They can also leverage partnerships to tailor customer insights, bootstrapping techniques or use proven pre-built analytical models, which have worked in environments with similar customer profiles or data sets.
AIM: How can Indian banks cultivate a more data-driven culture? What kind of infrastructure and skills do they need to invest in?
Kulkarni: Business leaders at banks need to advocate that every decision be supported with data-backed insights so that this approach permeates into the bank’s processes and culture. This must be supported by a structured change management program that embeds desired outcomes into routine banking processes.
Banks need to develop a comprehensive data strategy and make investments in an enterprise-wide data and analytics foundation, data governance, and management processes. A key element is a modern data and analytics platform that identifies and collates connected and contextual data from within and outside the bank and can convert data into insights for easy consumption. This platform needs to be supported by enabling architecture such as cloud-based accelerators and self-service tools. Where needed, there must be willingness to transition from legacy data systems to a scalable and modular data architecture, reconfigure processes and systems to support easy sharing of data across the organisation.
Investing in data leadership and cultivating data literacy throughout the organisation is equally important. Banks must build or hire for data, AI, analytical skills and related multi-disciplinary skills such as data visualisation, data storytelling and behavioural sciences.
AIM: Experts say that cloud adoption is vital to banks becoming more data-driven. Can you elaborate?
Kulkarni: Limited data storage and compute power are rendering on-premise data lakes and analytics environments sluggish and expensive. The cloud can help overcome this challenge since it is scalable and offers elastic storage and computation. It allows banks to integrate data from different sources and make it more accessible in real time. This can enable agile reports, more sophisticated analytical models and insights for faster decision-making.
Factors such as regulatory compliance related to data residency, customer data protection, security, and return on investments must be taken into consideration while envisaging a bank’s cloud adoption journey.