BNL - Legacy Modernization

Finance & Insurance

BNL realizes mainframe offloading via a Digital Integration Hub

Solutions:

Data Platform - Data Products

Technologies:

Confluent - Qlik

Business Summary

The bipolar structure of modern IT architectures is increasingly influencing the digital transformation paths of companies operating in the banking sector. On the one hand the mostly legacy back office systems, on the other hand the modern digital touchpoints that have to be constantly synchronized in as real time mode as possible in order to guarantee the user an experience that meets his or her needs as much as possible, putting a definitive end to the differences between services offered on-site and online.

In order to be able to guarantee its customers these omnichannel experiences, BNL launched a project in 2020 to build a platform capable of reading data from the legacy once in real time mode and then aggregating and consolidating it into continuously updated views that can be consumed by front-end applications.

The new platform built in collaboration with Quantyca has made it possible to make available and expose mainframe data (IBM Db2) to open banking systems, obtaining a constantly updated view of the customer (master data, account movements, products purchased, KYC, …), and drastically reducing the updating time of digital touchpoints, while at the same time lowering the workloads borne by the legacy and reducing the impacts, and therefore the costs, in terms of MIPS.

Challenges

The analysis of account movements and customer behavior is crucial for BNL to ensure positive experiences and to launch truly impactful initiatives. The high number of customers and transactions per day, coupled with the constant growth in the number of touchpoints, has led to an exponential increase in the costs and workload incurred by the mainframe in terms of MIPS.

Thanks to Quantyca, we have built a digital ecosystem available to our data scientists that enables the development and training of Machine Learning models, while at the same time being highly integrable with pre-existing banking systems. The need to expose mainframe data to open banking systems and digital ecosystems exposed the legacy to risks related to the loads required by digital channels, which could lead to problems with the stability and robustness of the infrastructure
Giovanni Cauzillo - Head of Data Intelligence Platform at BNL

The increase in digital channels, also fuelled by the needs due to the Covid-19 pandemic, has therefore required integration with an ever-increasing number of external applications, increasingly affecting the computational load of legacy systems. Data are in fact mainly extracted with batch integrations, forgoing real-time alignment and thus causing a proliferation of integrations.

Therefore, it became increasingly necessary to build a solution that could capture the data managed by the legacy system in real-time and bring it into a platform for building consistent and always-up-to-date views to be exposed to open banking systems, while maintaining a minimal impact of the latter on the mainframe and providing a single, secure and controlled access to the data.

Solution

The entire solution is based on a data integration architecture that decouples the legacy from the touchpoints, ensuring on the one hand the scalability and robustness requirements of the mainframe and on the other hand the speed of updating the digital channels, even cross-country.

We knew that in order to cope with the continuous growth of our digital ecosystems and the needs of our customers, we had to ensure greater reliability and robustness of our back office systems. To do this, we needed a system that could collect and expose data in a single format and at the same time reduce the demands on legacy systems by reducing workload
Giovanni Cauzillo - Head of Data Intelligence Platform at BNL

BNL, with the support of Quantyca, was able to create a digital platform that follows the Digital Integration Hub (DIH) paradigm and is able to offload data from the mainframe (IBM Db2) and consolidate it into integrated and constantly updated views that can be consumed by front-end applications.

Architecture

These are the main components that make up the platform:

A Change Data Capture (CDC) system implemented with Qlik Replicate and capable of capturing changes from the mainframe by offloading these changes in real-time
A Kafka-based Event Store for storing, integrating and processing events in real time
Kafka Streams applications that transform technical events from the mainframe into domain and business events consumable by consumers
A Replicator that allows topics to be easily and reliably replicated from one Kafka cluster to another
A consolidation layer for data exposure and consumption by front-end applications

Results

By collaborating with Quantyca, BNL is able to expose bank-wide customer data offering an up-to-date view of their habits, while mitigating the risks and workloads of legacy systems. The ability to consult constantly updated views allows BNL to offer its customers, through the immediate updating of its open banking systems and digital channels, high quality services in line with current and real needs.

Today, the platform processes around 4 million messages per day, with daily peaks of up to 10 million transactions, and with an average latency of less than 5 seconds, the impact suffered by the mainframe in terms of MIPS is completely negligible.