The challenge
Our client, a large Danish bank, recognised that it needed to transform its payments data estate: the complexity of the bank’s payments landscape meant that data was dispersed in multiple systems, and this made it challenging to obtain the accurate, end-to-end view needed to facilitate data analytics and Management Information (MI) use cases.
The data-consuming areas of the business – products and financial crime – were particularly impacted. Data had to be captured and interpreted from multiple sources in different structures, which was highly inefficient, costly and error-prone.
To address these challenges, the bank sought Icon’s expertise to create a canonical data model for a new centralised payments data store. This would enable a consistent representation of payments data across multiple payment data types, and the different markets in which the bank operates.
The approach
We leveraged our in-depth payments domain and data experience to define a logical canonical data model for payments and source systems mapping, based on ISO 20022 standard and adapted to the client's specific requirements. This involved:
- Conducting an analysis of the client landscape to understand the markets, payment types, related use cases and existing payment data stores
- Defining the conceptual model and the logical model with entities, relationships, and attributes. The model was expressed through an underlying metamodel with the ability to be visually represented without locking the client into a specific third-party modelling tool. This was key to delivering the flexibility to independently manage the model in the future
- Iteratively developing the model through regular sprints and ensuring the bank’s stakeholders were actively involved when it came to providing input and conducting reviews
- Mapping payments data from the source systems to the logical data model for key use cases: cross-border, and one core domestic market. Some parts of the data mapping were done collaboratively with the client team so that they could gain knowledge on the mapping process, to be able to conduct their own data mapping for other market use cases.
Solutions and benefits
The data model developed by Icon is playing a foundational role in the bank’s wider payments data transformation roadmap. By providing a mechanism to depict payment types across different markets in a standard, canonical form, it is easier to define the data products for consumption by different business areas.
Furthermore, these data products can now be used to enhance data analytics and deliver insights that can be used to create new service offerings, improve risk and compliance, and empower the bank to leverage data science and Artificial Intelligence (AI) technologies in future.
The payments data model also provided interfaces into other domains in the bank: customer and accounts. This enables the bank, in future, to easily link payments data with other domain data models to provide a wider enterprise view of the data.
The data mapping activity performed for the key use cases, provided assurance on the data model and enabled the bank to themselves perform further domestic market data mapping use cases.
Icon have certainly helped us progress on the data model, and we are thankful for your contribution and approach to the work.”
Our unique contribution
Icon leveraged its payments domain knowledge and wider data architecture expertise to apply data modelling best practices and processes, and define an approach that addressed key hurdles for the bank.
A particular element of this was the definition of a new tooling approach. Our client had not identified a tool for depicting and managing the data model, and so Icon devised a simple but effective approach that is enabling the bank to manage the model data and views. This approach has removed the need for the bank to use a licensed tool, while also enabling the flexibility to adapt the model as required in the future if the bank needs to import it to third-party data modelling products.
