img-woman-working-on-tablet-in-tech-office-with-servers-and-screens

The future of banking is hybrid

Integrating mainframe power with on-premises and cloud agility

Contact us

The future of banking is hybrid - integrating mainframe power with on-premise and cloud agility

The banking industry stands at a pivotal moment. While legacy mainframe systems like Cardlink and ASCCEND continue to be the backbone of reliable payment processing, the demand for AI-driven insights and personalised customer experiences is pushing financial institutions towards the cloud.

This paper explores a hybrid cloud strategy, demonstrating how banks can leverage the strengths of both their existing IBM Z and AS/400 infrastructure with modern on-premises and cloud platforms. We will dive into how this powerful combination can unlock a new wave of innovation – from enhanced fraud detection to hyper-personalised marketing – through  specific, high-impact use cases. Showing you how this hybrid approach can deliver significant business value and a clear return on investment.

img-full-man-sitting-at-a-table-in-cafe-using-smartphone-while-having-a-cup-of-coffee

Challenge: bridging the gap between legacy and innovation

Banks face a significant challenge: needing to figure out how to leverage the vast amounts of transactional data locked within their secure and reliable mainframe systems for modern, AI-powered applications. This is further complicated by the fact that moving  everything to the cloud is not always feasible or desirable. Especially given the security, reliability, and unparalleled processing power of mainframes for core transactions.

The key lies in a hybrid approach that seamlessly integrates the best of both worlds, creating a powerful synergy between stability and agility.

img-full-profile-view-of-young-programmer-working

Solution: a hybrid cloud strategy for AI-powered banking

A hybrid cloud strategy allows banks to keep core transactional systems on their secure and reliable on-premises infrastructure while leveraging the scalability and advanced analytic capabilities of the cloud. This approach provides the following benefits:

  • Modernise without disruption: continue to run mission-critical Cardlink or ASCCEND payment solutions on existing IBM Z and AS/400 platforms but integrate an on-premises/cloud data lake for AI and analytics

  • Unlock the power of data: securely and efficiently move your transactional data to a cloud data lake where it can be combined with other data sources to train powerful AI models

  • Drive innovation with AI: leverage advanced AI and machine learning services offered by cloud providers to build and deploy a wide range of applications. These can improve operational efficiency, reduce risk, and enhance customer experience
  • Maintain security and compliance: ensure the highest levels of security and compliance by keeping sensitive data on premises and leveraging the cloud for less sensitive workloads 

AI-powered use cases for the modern bank

By making transactional data from Cardlink and ASCCEND available in a cloud environment, banks can unlock a suite of transformative AI applications:

img-full-woman-ask-chatbot-on-laptop-looking-for-customer-support-online-screenshot-integration

Real-time fraud detection and prevention

Go beyond traditional rule-based systems, which are often slow to adapt to new fraud schemes. AI-powered systems continuously monitor transactions in real-time, analysing dozens of variables for each transaction – such as location, time, amount, and merchant type – against the customer's historical behaviour. These systems can reduce financial discrepancies by as much as 30%. By learning from historical data,  AI models spot subtle patterns and anomalies that indicate sophisticated fraud, significantly reducing financial losses and protecting the bank's reputation.

img-full-woman-hand-with-credit-card-swipe-through-terminal-sale-market-shopping-retail-concept

Advanced merchant services and payments

For merchant-acquiring banks, AI can be a powerful tool to personalise and optimise payment experiences. Using transaction data, AI models enable dynamic pricing based on volume and risk, create customised loyalty programs that reward specific spending behaviors, and provide merchants with tailored product recommendations to help them grow their business. This not only increases merchant satisfaction and retention but also opens up new revenue streams for the bank.

img-full-woman-looking-at-phone-screen-with-tea

Intelligent dispute resolution

The dispute resolution process is traditionally manual, time-consuming, and costly. AI can streamline and automate this entire workflow with dispute data, transaction details, and even unstructured documents like customer emails and merchant receipts, AI-powered systems can automatically categorise disputes, predict the likely outcome, and even recommend the optimal resolution. This leads to faster, more accurate dispute resolution, reduced operational costs, and a significantly improved customer experience. 

img-full-woman-using-a-digital-screen-table

Proactive digital collections

AI can transform the collections process from a reactive, often confrontational activity into a proactive, data-driven strategy.AI models analyse customer data (including payment history), transaction patterns, and even external economic indicators, to accurately predict which customers are at risk of delinquency. The bank is then able to tailor its collection strategies, offering personalised payment plans or other assistance, before an account becomes seriously overdue. This approach improves collection rates while preserving positive customer relationships. This approach improves collection rates while preserving positive customer relationships.

Marketing-smiling

Hyper-personalised marketing and product recommendations

In a competitive market, generic marketing campaigns are no longer effective. AI can track customer spending patterns and transaction history, allowing banks to deliver highly personalised marketing offers and product recommendations. For example, a customer who frequently makes international purchases could be automatically offered a foreign currency account or a credit card with no foreign transaction fees. This level of personalisation increases customer engagement, drives higher conversion rates, and builds lasting loyalty.

img-cut-ring-crop-sage-mint-young-asian-entrepreneur-trading-with-laptop-in-public-place

Dynamic credit risk scoring and management

Traditional credit risk models are often static and rely on a limited set of data points. AI models  can access a much wider range of data from Cardlink and ASCCEND, along with alternative data sources, to create more accurate and dynamic credit risk scores. These models can continuously learn and adapt, providing a real-time view of a customer's creditworthiness. This enables banks to make smarter lending decisions, reduce default rates, and more effectively manage the risk across their entire credit portfolio.

img-full-businesswomen-analysing-the-bar-chart-with-a-digital-tablet

Predictive customer churn analysis

Acquiring a new customer is far more expensive than retaining an existing one. AI analyses subtle changes in a customer's behaviour – such as a decline in transaction frequency, a shift in spending to a competitor, or a series of customer service inquiries – to identify customers who are at high risk of churning.  The bank can then proactively intervene with targeted retention offers, personalised service, or other incentives to keep valuable customers from leaving.

img-handling_fraud_in_the_digital-world

Enhanced anti-money laundering (AML) compliance

The penalties for AML non-compliance are severe. AI can significantly enhance a bank's AML efforts by evaluating vast amounts of transaction data to identify suspicious patterns that may indicate money laundering or other illicit activities. By flagging complex, multi-layered transactions and unusual fund movements that would be nearly impossible for a human analyst to detect, AI helps banks more effectively meet their regulatory obligations and avoid costly fines.

img-full-young-asian-woman-withdrawing-cash-money-at-the-atm

Data-driven ATM and branch optimisation

The physical presence of a bank, including its ATMs and branches, is still a critical part of the customer experience. Transaction data from Cardlink and ASCCEND can be combined with demographic and geographic data to optimise ATM placement, ensure optimal cash levels, and inform decisions about branch locations, operating hours, and staffing levels. This ensures that the bank's physical network is operating as efficiently and effectively as possible.

Leveraging the best of IBM and the cloud

We work with a suite of tools and solutions that enable you to build a robust and scalable hybrid on-premises and cloud infrastructure. IBM tools for data integration:

IBM DataGate

This solution seamlessly replicates your Cardlink VSAM (Virtual Storage Access Method) data into a watsonx.data lakehouse or other cloud data platforms. DataGate can be configured to replicate data to an on-premises instance IBM Db2 Warehouse, which can serve as the core of your private data lake. The principle remains the same –  real-time, low-latency replication from the mainframe to a powerful analytical database, all within your own data centre.

IBM Change Data Capture (CDC):

  1. Replicate your VSAM data in real time to a variety of cloud data lakes.
  2. On-prem functionality: IBM CDC is perhaps the most versatile tool for this purpose; it is designed to be destination-agnostic and captures changes from mainframe sources and can deliver them to a vast array of on-premises targets:

a) Hadoop Distributed File System (HDFS):  the foundational storage for on-premises data lakes built on platforms like Cloudera. CDC can write data directly to HDFS

b) Kafka:  an on-premises Kafka cluster can run as the central "event stream" for your data lake. CDC feeds real-time transactional changes into Kafka topics, which can then be consumed by various on-prem analytics engines like Spark or Flink

c) Traditional Databases: it can also replicate to on-premises relational databases that might be part of your analytics ecosystem

IBM Data Virtualization Manager (DVM):

DVM allows you to access your Cardlink VSAM data in real time from your data lake, with high zIIP offload capabilities to minimise mainframe processing costs.

Digital integration hub

Cache your transactional data on a z/OS cache and stream it to your data lake of choice via Kafka for real-time AI applications

Leading cloud platforms

We can help you integrate your on-premises data with the leading cloud platforms, including:

Amazon Web Services (AWS)

Leverage a wide range of services for data storage, machine learning, and data warehousing

Microsoft Azure

Build and deploy AI models, and migrate your AS/400 workloads for improved performance and cost savings

Google Cloud Platform (GCP)

Utilise powerful tools for data analytics, machine learning, and data warehousing

Snowflake

Create a single, unified platform for all your data needs, from warehousing to data science

Cloudera

Manage and analyse your data across hybrid and multi-cloud environments

Databricks

Build and deploy high-performance AI applications on a unified data and AI platform

The future is hybrid – contact us today

asian woman smiling leaning on desk with two men beside her, all looking at desktop screen
* Mandatory fields
Country
Marketing Optin

By submitting this form, I confirm that I have read and understand the terms of use from Worldline, and that Worldline will process my personal data as stated in the privacy policy.