How data and AI can transform financial services in Singapore

THE pandemic and its effects have thrust companies, especially financial services institutions (FSIs), into a new reality where digital transformation is no longer the most optimal strategy.

Driven by factors such as rising customer expectations of personalised services and the need to automate back-office processes, FSIs must now go beyond digitalisation to thrive in the post-pandemic world.

Today, some FSIs are still relying on manual processing and legacy systems to gather and store important customer data. This becomes an issue as the speed at which data needs to be processed today is different from the past. Beyond daily operations, FSIs are also struggling to cope with an increase in back-end processes such as data analysis, assessments, and strategic planning. With the amount of data and knowledge that must be processed, FSIs need to start leveraging AI.

So, how can FSIs take the first step in leveraging the power of AI? It starts with building a collaborative data platform.

The data building blocks

Many financial organisations currently can store and clean data, pull off reports and ad hoc queries for insights from historical information. However, as data volumes increase, they can get stored in the wrong places leading to data duplication and inconsistencies. This will hinder efforts to tap data analytics and AI to capture important insights, such as fraud patterns, customer behaviour and investment intelligence.

To accelerate innovation and transformation, organisations should be looking at the wider potential of data. Analytical approaches - under-utilised by organisations - can yield tremendous value, such as data exploration (why did something happen), predictive modelling (what will happen), and prescriptive analytics (how can we make something happen).

To maximise the potential value of advanced analytical approaches, they must be underpinned by a robust modern data architecture that will become even more important as open banking increases options for consumers.

Currently there are multiple architecture options for efficiently storing, cleaning, and analysing data - data warehouse, data lake and data lakehouse. Both data warehouse and data lake have their own strengths and weaknesses when it comes to what data can be stored and how the data can be analysed. However, data lakehouse marries the best of both architectures, emerging as the necessary data structure for organisations to draw out crucial insights.

Data and AI making an impact

Getting the right data architecture in place, such as a lakehouse, is the crucial first step for any organisation looking to reap the rewards of using data and AI. Here are key areas where data and AI can transform FSIs for the better:

• Personalisation

Data and AI can help create a more personalised customer experience and move financial firms away from product centricity towards customer centricity. Continuous intelligence - marrying event-driven decision making and historical context - ensures completely personalised interactions with customers based on the analysis of millions of unique data points every second from multiple sources. By innovating looking at customer insights first, products can be aligned with real-time behaviours and needs.

• Fraud detection

Fraud detection at scale is no easy feat, particularly as data volumes increase and online fraudsters switch up deviant tactics to avoid detection. Having data in one place helps with scale and visibility, and also provide an easy framework for sourcing out fraud at its roots.

Organisations can build a fraud detection data pipeline to visualise the data in real time. This allows more flexibility than setting rules on how fraudsters behave and mapping this against a subset of data to detect possible fraud cases. For example, Singapore's Anti-Scam Centre (ASC), set up in partnership with leading Singapore FSIs, has enabled local banks to leverage machine learning and analytics to monitor massive amounts of banking data and more easily spot suspicious accounts and fraudulent activity.

• Risk management

A modern, agile risk management practice is the way forward to managing and responding to market and economic volatility. While historical data and aggregated risk models run the risk of obsoletion, data and AI enable delivery of scalable, real-time insights allowing FSIs to tackle threats efficiently. Under an initiative known as Veritas, a 25-member consortium of leading banks and e-commerce giants have come together to evaluate their AI and data analytics-driven solutions against the principles of fairness, ethics, accountability, and transparency.

The future is open

An open, simple and collaborative approach to data and AI will propel the financial services industry forward in many ways and accelerate innovation. It may be a heavily regulated industry, but the richness of customer data and its velocity bring many opportunities for positive change and disruption, all the while keeping customer convenience and security at the heart of business growth.

As FSIs push ahead to keep up in the increasingly competitive industry, ultimately, embracing "openness" and "simplicity" will liberate them from the throes of an outdated system and propel them into a space of innovative, lower-cost and consumer-centric banking system.

• The writer is global industry leader of financial services and sustainability at Databricks