What Does It Take to Become An AI Bank?

AI is taking banking by storm to such an extent that in the future there will only be 2 kinds of banks:

—     Banks that make use of AI

—     Banks that natively integrate AI

The difference between the two might not seem that big in the first place, however in reality the gap is huge. The reason is that the banks’ ability to integrate AI functionality will be much more critical than their ability to offer it in the first place. In other words, if AI is only superficially attached to the banking operations, then it cannot become a real competitive edge and, in turn, the organization cannot reap the benefits that come with a full AI digital transformation.

What does this mean for banks?

AI is essentially shortening the banking transformation cycle by a very high degree by converting the long digital transformation journeys (ask any banking executive what they have been working on during the past years and chances are that digital transformation will be among their top priorities) to a complete metamorphosis that must happen in a fraction of the time that such projects used to take.

One of the most impressive examples of such journeys is coming from Siam Commercial Bank (SCB), Thailand’s second largest commercial bank that has literally become a benchmark for banking digital transformation in just a few years.

The pandemic gave SCB the trigger they needed to take the decision to transform. Not just transform but become an AI-first bank.

Within a few years SCB managed to make AI the core of their operations and strategy and:

—     attract 250,000 new customers

—     lend up to 1,2 billion Baht in a short amount of time

—     have 84% of their user base (14.3 million users) going digital

There are many aspects of how they have managed to pull it off, cultural, operational, technical, but I would like to particularly focus on one of the main enablers:

A strategic approach to leverage cloud computing as the basis for building a core digital banking platform and the decision to do so via a cloud-native provider.

The infrastructure is Huawei’s Cloud Native 2.0 platform, which is based on the concept of a dual strategy using AI to optimize cloud infrastructure ("AI for Cloud"), while simultaneously leveraging cloud resources to enhance AI development and deployment ("Cloud for AI").

Here are some of the key elements of this approach that helped SCB to radically transform:

—     A big data and AI convergence set-up aimed at solving governance problems caused by data silos between lakes, warehouses, and AI platforms.

 —     A partner approach employed to launch smart process robots, integrate AI, robotic process automation (RPA), and low-code capabilities. This intelligent automation is a big efficiency booster for financial services organizations. A good example in SCB’s case is Huawei Cloud's partner Sunline, a leading banking software and technology services company that built an innovative digital loan application platform based on Huawei Cloud's solution. The platform used AI technology to determine the credit worthiness of customers, then allocate appropriate credit limits and finally make relevant proposals on products and services.

 —     Low latency capabilities, which means allowing a computer network to process a very high volume of data messages with minimal delay. For SCB this translated into users being able to access their platform within 0.1s.

 —     The Pangu graph model, which uses groundbreaking graph network fusion to improve the accuracy of financial exception identification to 90%, spotting financial exceptions in minutes and greatly enhancing enterprise risk identification.

 —     Capabilities such as unified storage and metadata, so that a copy of data can be used between multiple engines quickly and consistently.

 —     High scalability that allows client banks to optimize cost and resources by a dynamic allocation to optimize usage between peak and off-peak hours.

The game has radically changed for banks in a very short amount of time. Not everyone can transform in months, but there are a couple of common requirements to be able to successfully adapt in the new environment: 1) being able to leverage an “Infrastructure as a Service” approach from the right partners 2) do so in a set-up that brings together AI, data and the cloud in a native way.

Link for additional info on Huawei Cloud and the SCB use case: https://www.huaweicloud.com/intl/en-us/cases/scb.html