Core banking packaged software – an industry at a crossroads

In 2017, banking technology expert and former Digital CIO at Lloyds Banking Group, Jon Webster, stated that what the banking industry really wants and needs is a “Kubernetes of Banking” – a core technology platform so compelling and cost-effective that banks have no choice but to migrate to it from their existing core systems.

AI will lead to the next evolution of banking software but will not be the death knell of it

The Kubernetes of Banking would be open, cloud-native, intelligent, and designed for seamless scale and agility. It would enable continuous innovation and lower costs by 10x. It would be the foundation for a platform strategy and enable its creator to orchestrate a digital ecosystem around clients, partners, and third-party developers.

Core banking vendors – the battle to survive

Almost ten years on, not only does such an intelligent platform in core banking not exist, but questions are being raised as to the viability of the core banking software industry itself. No software vendor, whether incumbent or a neo-disrupter, has succeeded in building the Kubernetes of Banking so far.

Furthermore, with the advent of AI – which lowers switching costs and barriers to entry, increases banks’ ability to modernise themselves, and could even spell the end of software itself via AI agents – are core banking software vendors facing an existential threat? How long can they continue to rely on the success and stickiness of their installed base, and on an addressable market predicted to grow steadily based on historic third-party banking spend?

The vendor landscape today

The global core banking packaged software industry has undergone a seismic change in the past decade. From a handful of strong incumbent vendors, the landscape has evolved into a complex web of players, all vying to differentiate in a crowded market. These include cloud-native vendors, BaaS players, and challenger banks monetising their technology platforms.

Although comprehensive functionality across multiple jurisdictions and proven credentials on legacy migration, performance, and scale remain competitive advantages for incumbents, the neo-vendors have set new standards in terms of agility and cost. They have made banks aware of the “art of the possible” – 24-hour deployment, less than 90 days to go live, try before you buy (sandboxes), choice and flexibility of which components to buy so that banks can slice and dice the scope of their digital transformations any way they like, with minimum disruption to the business and quick time to value.

On the flip side, the neo-vendors are yet to scale and showcase migration from legacy, even those that have been around for more than a decade.

Both types of vendors need a serious injection of funds to achieve the market-making dominance of a Kubernetes of Banking. Indeed, I believe existing vendors need a reinvention from the ground up to bring their product and sales organisations up to the technology standards of the third decade of the 21^st century, while the neo-vendors need the time and money to build the functionality required to demonstrate evidence of large-scale migration and performance.

However, core banking ISVs struggle to make the necessary investment because focusing on long-term growth is difficult when in the thrall of quarterly margin pressures and performance targets if a listed company, or when in the pincer grip of investors with short exit timeframes if privately held.

Build rather than buy?

Theoretically, AI has the potential to reduce banks’ appetite to buy and to help them choose to build themselves instead. If banks were to deploy AI in their own internal development shops, this would disintermediate third-party vendors. In development and maintenance, AI and GenAI could produce initial code, auto-fill or predict next steps for bank coders, and help them debug. It could transform legacy code into modern cloud systems and produce synthetic testing data and run automated tests.

However, as of today, AI technology lacks the context and is limited to specific tasks and simple blocks of code only, rather than a complex chain. Also, the vast majority of banks, barring the mega players, do not possess the breadth and depth of data required to train models in the way a software vendor with lots of banking clients could do. In this context, the addressable market for third-party banking software remains secure for the foreseeable future.

Smart agents and dumb applications

Now we come to Microsoft CEO Satya Nadella’s recent comments about AI potentially killing off SaaS or business applications. In an appearance on the BG2 podcast, Nadella suggests that AI-powered agents will soon replace traditional software models, and business logic will move from the applications into the AI layer of an intelligent platform, thereby in effect reducing these applications to mere databases. These integrated platforms would unify disparate point solutions into a cohesive ecosystem, making workflows faster and smarter. All the decision-making, automation, and orchestration would sit with the AI agents rather than the business applications.

However, apart from the fact that it may not be cost-effective or practicable for this to happen in any industry, banking is a very different beast. Workflows in banking are complex, regulated, fragmented, and not fully automated. The industry has not even embraced SaaS yet, let alone AI. How would agentic AI deal with complex core banking business rules designed for retail, corporate, and private banking across multiple jurisdictions?

The Kubernetes of Banking platform, with its modern APIs, reusable tools and components, cybersecurity, and rich data, is paradoxically a pre-requisite for Satya Nadella’s prediction to come true. Yet, AI could be the catalyst for the evolution of just such a platform. Whoever is able to use AI to build the most functionally rich and technologically advanced front-to-back banking platform and extract real-time transactional data from mission critical applications, behavioural data from channels, and unstructured data from within and outside the bank, could unleash AI and AI agents onto this data to become the winner in this industry.

A history of evolution and not revolution

This is difficult but not altogether impossible. Time and again new technologies have been predicted to kill off banking software, but core banking has defied the hype and embraced the best of them. It became cloud-native and hence hyper-scalable, secure, open, and cheaper and easier to implement and run. It used blockchain where it made sense – in payments and trade finance. Microservices were the next big thing – everyone said monolithic core banking software was never going to survive unless it broke up into a thousand microservices. The smartest banking architects came up with the right solution: the Goldilocks principle of banking capabilities broken into domains at a level of granularity not too big and not too small, but one that is appropriate for the average bank from a business and future operability perspective.

AI will lead to the next evolution of banking software but will not be the death knell of it. Banking technologists will ride the AI wave and not get carried away by the hype, just as they have always done with disruptive technologies. The more likely scenario is that core banking applications will embed AI into their products and architecture and become more integrated, intelligent, and autonomous. The aim would be to use AI to deliver tangible business value to banks, rather than technology for technology’s sake. The first mover in this space, with the long-term vision and ambition, would then be in pole position to become the Kubernetes of Banking.

About the author

Kanika Hope is a senior fintech leader with 30+ years of banking technology experience at organisations such as Temenos, SAP, McKinsey and General Electric. She brings deep industry expertise in market and product strategy, value-based selling and business development in banking application software, as well as functional expertise in IT strategy and transformation in financial services. All opinions are her own – feel free to debate and comment below!