Core banking integration: Models, approaches, benefits, challenges and best practices

Key takeaways

• Core banking integration enables internal and external applications to exchange data with the core.
• Integrating core banking with other systems helps automate workflows, unify data and analytics, leverage third-party capabilities, and comply with open banking regulations.
• Digital assets are moving from pilot to product: stablecoins and tokenized deposits are gaining ground in cross-border payments.
• Core banking integration can rely on APIs and middleware, use event-driven or message-driven models, and enable coexistence and parallel core modernization.
• Core integration approaches include point-to-point, hub-and-spoke and Enterprise Service Bus (ESB) integration.

If data itself is the new oil, then integrations are the pipelines that get it to where it’s needed, when it’s needed. Without those high-speed, reliable pipelines, it doesn’t matter how much oil you have or how valuable it is — it’ll just sit there, until it’s too old to be useful anymore.

Core banking systems hold, perhaps, the most valuable data within a given financial institution.

Yes, making it readily available is the cornerstone of compliance with open banking regulations, now in place in 59 jurisdictions worldwide. But compliance isn’t the only reason to give your core banking integration some thought. Seamless data flows can also power AI initiatives, automate workflows and enable new revenue streams.

With the average age of a core banking system around 20 years, however, integrating it with other systems is no cakewalk. Here’s why — and how to succeed despite these challenges.

To understand core banking system integration processes, you first need to get a full grasp of its key components and technologies.

Core integration building blocks

A core banking integration involves the following components:

• API gateways. An API is a protocol that enables applications to exchange data securely. Modern integrations typically rely on RESTful and SOAP APIs.
• Middleware solutions. A middleware layer uses an Enterprise Service Bus (ESB) to connect applications that can’t be connected through APIs. It’s a centralized communication system that standardizes and routes messages.
• Authentication layer. This layer ensures the application and the user have the right to access the desired data. It typically relies on OAuth 2.0 and OpenID Connect protocols. 
• Security layer. In addition to secure access, you should also ensure that sensitive data is stored and transmitted securely. That involves encryption, tokenization and compliance with regulations (PCI DSS, PSD2, etc.).
• ETL/ELT pipelines. Extract, Transform, Load/Extract, Load, Transform processes clean, standardize and transfer data between systems. They’re crucial for preserving data integrity and quality.
• Regulatory and compliance module. This component maintains audit trails and automates checks to prevent compliance breaches.

Technologies behind modern core integration

We’ve already mentioned multiple technologies that power core integration: APIs, middleware solutions and ETL/ELT pipelines. While they remain staples of integration projects, several new players have entered the field.

Low-code and no-code development tools, for example, speed up custom API development with drag-and-drop interfaces and pre-built software components.

AI, in turn, has emerged as a new way to automate API management. For example, AI-powered API management solutions can:

• Detect anomalous behavior to prevent breaches
• Dynamically allocate resources based on traffic patterns
• Identify integration bottlenecks

In addition to that, AI can also streamline ELT/ETL processes by speeding up data cleaning and structuring for large datasets.

Due to their nature, core banking systems hold and process a wealth of account, customer and other data. All that data can enhance other systems. For example:

In the context of integration, the term “touchpoints” can refer to two things. On the business side, they represent the specific applications connected during integration. For example, a core banking system and a mobile banking app can be two touchpoints.

On a more technical level, touchpoints refer to the exact data or functionality exposed during an exchange. For example, if the core banking software is integrated with a payment system, touchpoints on the core banking side can include:

• Customer account balance
• Customer/account validation
• Payment initiation request

At the same time, the payment system can expose the following data to the core banking software:

• FX exchange rates
• External customer data
• Payment status updates

Integrations are versatile; they can power everything from marketing automation to digital banking channels. Here are the most common use cases for core banking integration:

During core systems integration, banking and third-party systems can be connected with the core banking software in many ways. Let’s break down the most common architecture models.

API-centric integration models

The de facto standard for modern integrations, APIs are popular thanks to their simplicity and security. That said, APIs come in many flavors, enabling different kinds of integration:

• Request-response integration. One system initiates the action, the target system processes it and returns the data.
• Event-driven integration. The system “listens” for changes in the other system’s state or data.

Event-driven and message-driven models

Event-driven and message-driven models are communication patterns that essentially determine how often data will be sent between systems

• Event-driven architecture (EDA) uses events to detect changes and notify the other system about them. It enables real-time data exchange, which is a must for real-time fraud detection, automated alerts and instant money transfers.
• Message-driven architecture (MDA) uses messages to exchange data between systems synchronously or asynchronously. It supports high-volume messages but introduces latency, making it less suitable for real-time data flows.

Coexistence and parallel core models

These two models represent possible solutions to a specific conundrum that arises during core banking modernization. As core banking systems are mission-critical and migrating to a new one takes three years, organizations typically opt for component-based replacement. In the meantime, new and legacy cores need to coexist.

That’s exactly what coexistence means: using integrations to keep the new and legacy components running together as one. That can be achieved through adapters, middleware and message- or event-driven management.

Where coexistence is a temporary (yet necessary) solution, parallel core is its more permanent sibling. Under this approach, the organization retains the legacy core but integrates a new one to close capability gaps. It’s also known as “building on the side” or “on top”.

On the macro level, there are three ways you can approach integration:

• Point-to-point integration. Under this model, every system is connected to the core banking application directly. It’s easy to implement but may create a hard-to-manage mishmash of integrations.
• Hub-and-spoke integration. To avoid the messiness of point-to-point integrations, you can connect all systems to a central hub (message broker) with an integration engine. This approach improves scalability but may create a single point of failure for all integrations.
• Enterprise Service Bus (ESB) integration. ESB integration also relies on a central layer, the messaging backbone. In contrast, each connected system has its own integration engine. This approach improves scalability even further, but the tradeoff is more complex troubleshooting and maintenance.

As a trusted core banking integration partner, DXC knows like no one else: integrations might seem simple on the surface, but they’re anything but simple. Years or even decades of customizing the core turn every integration project into an investigation.

Here’s a brief excerpt from the integration playbook we rely on:

• Embed compliance into technical requirements from the get-go (e.g., PSD2, PCI DSS, GDPR, etc.)
• Develop a unified data strategy to maintain data quality and integrity; otherwise, you risk the “garbage in, garbage out” scenario
• Always conduct sandbox testing before rolling out an integration to production
• Choose carefully between real-time data exchanges and batch processing

We’ve already mentioned the core customization quagmire above, but it’s far from the only challenge you may encounter on the way.

Depending on the core banking system in question, you may also run into issues stemming from legacy tech stacks. For example, compatibility and interoperability limitations can take certain types of API integrations off the table.

Of course, there’s also the challenge of cleaning, standardizing and validating all incoming and outgoing data and ensuring its quality. Implementing a unified solution to centralize data is one approach we highly recommend to our banking clients.

Finally, there are challenges associated with third-party integration platforms. While they can simplify the whole ordeal, you may risk vendor lock-in. Using such a platform also adds a third-party ICT risk.

In addition to the best practices we’ve mentioned above, there are three areas where getting integration right is crucial: APIs, regulatory compliance and AI use.

API-first integration: Best practices

Let’s be clear: not every integration project will necessarily rely on APIs. Yet, there’s no denying that APIs are the de facto modern standard in core integration for digital banking.

If your integration project involves using APIs, make sure to:

• Keep exchanges secure. Implement robust authentication and authorization methods (e.g., OAuth 2.0, API keys); encrypt data in transit.
• Audit APIs. Regularly check your APIs for security vulnerabilities, compliance breaches and performance issues.
• Put a governance framework in place. Create a set of standards and policies to dictate how you’ll develop, deploy, use, monitor and manage APIs.

Designing for regulatory compliance

Compliance risks are always at the forefront of banking executives’ minds, and for a good reason. Those risks don’t miraculously go away during core banking integration.

So, before you even select the integration method, outline regulatory requirements first. Pay attention to:

• Regulatory reporting: Audit trails, data accuracy, reporting automation
• Access controls: KYC checks, strong authentication, segregation of duties, etc.
• Financial crime prevention: Automated AML and CTF compliance, real-time detection
• Payment protection: Compliance with PSD2, PCI DSS, etc.
• Data privacy: Data encryption, anonymization if possible, consent management
• Third-party risks: Compliance with regulations like DORA in the EU

Using AI to enhance integration

AI is a potent tool that can streamline virtually every step of integration engineering and testing. For example, AI-assisted development tools can generate code suggestions, automate bug and vulnerability detection, generate test cases and optimize tests themselves.

That said, there is one crucial point to make about AI use here: AI can’t replace integration engineers, but it can make their jobs easier and faster. That’s why our AI-augmented modernization-as-a-service offering involves human-in-the-loop validation at every crucial step, for example.

Every bank has a unique software estate and a specific set of goals and needs. So, a cookie-cutter approach never works.

To chart the path for integrating your core banking system, first get the lay of the land — i.e., analyze your core banking system’s integration capabilities and limitations. Then, identify the systems you’ll integrate it with; however, ensure that every integration generates business value.

With that done, consider:

• Types of integration (on-prem to cloud, cloud to cloud, etc.)
• Regulatory compliance requirements (privacy, security, logging and monitoring)
• Integration complexity and orchestration needs
• Required update frequency (real-time vs batch processing)
• Data volumes and performance requirements
• Long-term scalability requirements
• Data quality and cleaning, standardization and validation requirements

The list above is just a snapshot of the key factors that will impact your integration architecture, approach and toolkit. The full list is much longer, and assessing each factor requires both technical acumen and experience with banking software and business processes.

That’s why turning to a core banking integration partner is often the right way for banks. But choosing one can be admittedly difficult. To make a well-weighed choice, we advise you to assess each vendor’s:

• Experience with legacy or next-generation core banking systems (depending on which one you have)
• Expertise in working with applicable regulatory requirements, including open banking
• Technical expertise in a wide range of integration architectures, patterns and approaches
• AI-augmented software engineering capabilities to speed up implementation

Over the past four decades, DXC has built an unparalleled expertise in banking software, including core banking systems. From offering our own Hogan core banking system to providing modernization-as-a-service, DXC is a trusted expert in implementing, changing, migrating, modernizing, maintaining and integrating core banking systems.

For example, our Hogan API solution enables banks to leverage open RESTful APIs to and from the core banking system, and the point-and-click interface makes creating API fast and easy. Our CoreIgnite platform, in turn, helps banks run parallel cores and orchestrate a myriad of integrations, all while maintaining data quality, security and compliance.

As for concrete examples of the value DXC brings to the table, discover how we integrated a mutual bank’s core banking system with Microsoft Dynamics 365 for real-time customer insights.

While consumer expectations and open banking regulations mandate connectivity, there is an opportunity to be seized here, too. Integrations break silos and centralize data, enabling consolidated insights. They power automation, creating operational efficiencies and cutting costs.

Connectivity is the new normal. Banks have to adapt to it — or they’ll risk losing their competitive edge.