Read the article at Finance Derivative

In the start up world, everyone is looking to do more with less, and the banking sector is no exception.  Challenger banks are testing the traditional monolithic high street institutions to find more agile ways of working and Al can be an attractive prospect.  But deploying such an advanced technology is not a plug-and-play scenario.

The reality of the situation is that, for many banks, we’re still at an interim stage when it comes to Al. There are examples of incredibly sophisticated systems being used in fraud detection and loan authentication, but it would be a push to say this was the norm.

The short answer is that AI is becoming a matter of necessity – from all sorts of perspectives – cost, risk, and market competitiveness forming just some.

A new generation of consumers expects services to be available 24/7, to be intuitive, instantaneous, and catered to them. For banks, this means providing services that cater to an individual’s spending patterns, device preferences, and current life situation. It’s a tall order for businesses to provide manually, but all the data on customers is there to provide those kinds of services – banks just need to find a way to access it.

Successfully accessing, and then analysing, data will lead to more streamlined services for customers and quicker time to insights for businesses means quicker service delivery to customers. Banks can also capitalise on their insights around customer spending patterns to provide tailored recommendations on financial wellbeing to clients – boosting their customer experience.

At the same time, legislation changes are adding additional complexity. As part of the shift towards these hyper-personalised services, the Second Payment Services Directive (PSD2) mandates that customers are able to request that third party providers can access their banking data to provide new services for the customer. The upshot of this is that banks will need to find a way to categorise, group and structure that data so that other services can plug in on request. Achieving this requires automation of the data structure to pull together, and analytics to determine which information is relevant for the required service.

And then there is the need for compliance with legislations like the IFRS 9 and GDPR. Both of these require the creation of “single points of truth;” the first on financial assets and liabilities, and the second on customer personal data. Both of these are exceptionally onerous tasks, if not outright impossible, to complete manually, and banks need some sort of system that can intelligently pull together all the insights into one, clearly representative, report of compliance.

Stepping stones to AI

If AI is a necessity for consumer personalisation and legal compliance, why aren’t banks everywhere simply investing and deploying it? It has emerged that, while there is great appetite for more analytics, more AI and more insight in general, all of these require banks to overcome significant legacy infrastructure hurdles. You can’t plug Al into an old system like a USB drive and expect it to churn out the results you want; instead, you need to get your data into a more searchable, agile framework before you can add AI over the top.

To do this, banks need to ensure they are predominantly focusing on stepping stone technologies – such as data warehouse automation. Data warehouse automation can provide a vital bridge between the legacy infrastructure that is holding many organisations back, and the future of cloud-based, data agility, by automating a lot of the manual, time consuming migration tasks.

This is particularly useful for traditional banks, who have substantial legacy infrastructure environments.  These banks tend to have a very large number of old mainframe systems, with little or no modern analytics capabilities – something that they are, or should be, seeking to change as quickly as possible. The more diverse the IT infrastructure landscape, and the more sprawling the legacy and modern systems, the greater the need is to find a way to streamline all this and deploy more sophisticated, quicker, analytics capabilities.

Data warehouse automation can help with this – streamlining and accelerating the migration process. In addition, correctly deployed automation can reduce many of the different potential risks that come alongside modernisation: risk of error, risk of doing things slowly, risk of human oversights. And, in addition, the cost savings of automating data ingestion processes with data warehouse automation can allow banks to be more competitive, and more innovative. In the greater scheme of things, where the monolithic high street banks are feeling the push from newer challenger upstarts, it’s never been more important to find ways to become more agile with data and insight technologies.

IT infrastructure, and particularly analytics and AI capabilities, are going to evolve significantly in the next few years. At the same time legal requirements like the GDPR, PSD2 and IFRS 9 change frequently –and will continue to do so. In order to keep pace with these changes, banks will require increasingly comprehensive data ingestion, ways to manage their data landscape, and faster, cheaper time to insight/value systems. During this transition, data warehouse automation will be a critical stepping stone between current legacy environments, and a bank’s AI-driven future.