Martin James, Regional Vice President, Northern Europe at DataStax
Artificial Intelligence offers huge opportunities to banks, but how can they use the data they hold more effectively?
Artificial Intelligence (AI) is getting huge amounts of investment from enterprises – according to IDC, spending on AI and cognitive computing went up by more than 50 percent in twelve months from 2017 to 2018 to an estimated $19.1billion worldwide.
The financial services sector is currently leading on AI investments, with the emphasis on real-time transaction analysis and smart fraud detection through to algorithmic trading and AI-managed funds. Banks and financial institutions are beginning to see the inherent potential of AI, especially when it comes to attracting new business by delivering better, more personalised customer services.
WANT TO BUILD A FINANCIAL EMPIRE?
Subscribe to the Global Banking & Finance Review Newsletter for FREE Get Access to Exclusive Reports to Save Time & Money
By using this form you agree with the storage and handling of your data by this website. We Will Not Spam, Rent, or Sell Your Information.
Understanding the different technologies – and why they matter to banking
The real attraction of AI is that it goes far beyond the boundaries of traditional computing into the realms of to understanding and extracting value from the massive amounts of data banks possess.
For example, Macquarie Bank talks about creating both oversight and foresight as part of the bank’s AI-based digital banking platform. This includes providing information to individual customers on their accounts, and using analytics to give the bank more insight into customers. AI is not only about smart automation – it’s about the insights that come from analysing data.
Machine learning enables computers to ingest data in order to ‘learn’ how to progressively improve the performance of a task. With this ability to sift and make sense of large amounts of information, machine learning in the banking industry has been able to reduce the time spent processing credit agreements, for example, from hundreds of thousands of man hours to seconds. With hundreds of different tasks across banks that can benefit from automation, the potential for AI to reduce costs is huge.
However, the majority of banking interactions rely on context. Customers may have multiple accounts and services with their bank – however, many banks struggle to build up that complete picture of the customer. For customers, the issue of feeling that their bank does not get them or understand their needs is one of the principal reasons for dissatisfaction. Yet the reason why banks have failed to build up single customer views is not due to a lack of data. Instead, the problem is how to manage this data effectively at scale and across different silos.
Using data effectively – based on new technologies like graph analytics – is an effective first step to delivering more joined up services. With this improved understanding of customers, it is then possible to look at how AI can be applied to understand customer behaviour and expectations. This can now be applied to customers with accounts across multiple banks as well – the advent of PSD2 and Open Banking means that banks have to share account data with each other and with third parties, which should make it easier to build up this picture of each customer based on their real-world activities and history.
It’s here where AI becomes so attractive. By recognising patterns in customer behaviour based on the data that they create, banks can make more recommendations to customers that actually fit their needs. By looking at what is taking place for that customer at that point in time – rather than thinking about group behaviour and what might be useful – banks can target their service offers more efficiently and stand a better chance of success. Secondly, automation and AI can spot services that customers may benefit from and make the process to use them easier.
Customers are always more likely to return to a service that’s easy to use and genuinely helpful. The benefits of AI are obvious, but if banks set themselves on a path to further adopt it, they will also need to pay greater attention to the underlying technologies that enable it.
What steps are necessary to improve the customer experience, and how are banks able to get there?
With the growth in mobile and web-based applications, there’s never been so much choice for customers who want to access banking services through their desktop, phone or tablet, whenever and wherever. In today’s fast-moving ‘right-now’ economy, customers have come to expect an immediate, integrated and seamless user experience, and service providers that don’t deliver get dropped.
Often, the most complex part of AI is not developing the algorithm but managing the data layer. For banks exploring exciting new AI use cases, their deployment strategy can become just as much about data management as AI. How do they extract the data they need out of traditional core banking applications in order to successfully deploy these AI technologies?
Across most businesses, data has historically been gathered – and is often still gathered – in an unstructured way. In banking, customer information can be spread across multiple accounts and many siloes. Getting the most out of data in order to gain insights into customer behaviour and to give more personalised customer experiences can be difficult when using legacy database technology to understand complex data relationships.
Platforms with capabilities to build a Single Customer View (SCV) from various sources of internal data are able to more easily identify individual customers. They not only help with compliance and privacy, but also give context to customers’ relationships with the company meaning personalised CX can be delivered.
In order to support new service design and delivery, data must always be available. Downtime can cost banks millions of pounds, so it’s important to ensure critical applications have always-on access to data. A solid data structure that eliminates any single point of failure in any kind of multi-cloud and/or on-premise architecture is essential.
Real-time access to data is also important. Combined with internal customer records, real-time capabilities and transaction analysis can give instant insights into customer behaviour, telling banks when someone might be receptive to new product marketing or when they might be unhappy with a service. Theis information give banks a competitive edge when it comes to delivering personalised recommendations, support and even innovative new services.
Analytics are important for helping to meet the needs of the today’s hyper-connected ‘right-now’ customer. In banking, the ability to respond quickly to issues such as errors and fraud is essential, and this means understanding the data landscape moment-to-moment. Without this real-time insight, customer experience will be poorer and issues will be flagged too late. In banking, a service offer that is made five days or five minutes late will be treated in the same way – with disdain.
For banks that operate on a global scale, it’s critical that customers can immediately obtain their information from whatever location they’re based in. But that means the data must be accessible anywhere too. In order to achieve this, it’s necessary to use a data platform that can manage widely distributed data and that replicates and synchronises data across whatever server infrastructure is being used, whether that be hosted with a cloud provider, on a bank’s own data centre or combinations thereof.
Today, the amount of data, the types of data and the speed at which data can be created are growing exponentially. Traditional systems are sometimes unable to process massive increases in data and are not always built to be adaptive. Data platforms must, therefore, be able to manage data with continuous and predictable scalability.
Data platforms with built-in analytics can also help banks plan for future architecture, understanding where they need to build in flexibility and scalability. Data management is a fundamental layer of AI technology, so getting the key architectural decisions right today will help companies wanting to embrace AI to prepare for tomorrow.
Useful source materials: