Open Banking is gaining momentum with many state regulators pushing for sharing customer data - with consent - to fuel innovation, improve efficiency and enhance customer experience. The Competitive Market Authority (CMA) in the UK, the European Parliament’s revised Payment Services Directive (PSD2), Australian Competitor and Consumer Commission (ACCC) are some of the regulators driving open banking in their geographies.
In the realm of open banking, data fabric is an approach that enables data management across the enterprise. This ensures consistent and secure storage of data, access and management. Banks can leverage the data fabric approach in product and service innovation by applying machine learning and artificial intelligence for newer business models and superior customer experience. They can also use the approach to serve open banking requirements, which in turn can eliminate the need for serving open banking APIs from their legacy core banking systems.
Open Banking: Many Opportunities, Some Considerations provides a view on regulatory adoption, our strategy and opportunities for the banks. As open banking mandates sharing the customer data about accounts, products, transaction history - and in some cases, also enables payment initiations - banks traditionally service these needs from their core banking systems.
Why do we need Data Fabric in Open Banking? How will it work?
Banks are looking at efficient ways to enable data for open APIs that originate from third parties, without compromising security and privacy requirements. They are also focusing on consuming data from third parties to get customer insights by combining the information from enterprise systems and use this to build innovative, new products or services. However, with most enterprises an increase in API invocation may lead to increased costs with huge legacy dependencies (e.g., core mainframe systems) for servicing the APIs. This is where the open banking data fabric can help by providing a flexible means for ingesting data from enterprise and external systems, storing data securely, generating insights and providing access through data services (APIs) while maintaining cost efficiency.
Several enterprises are looking to leverage the existing investments on data estate. The firms are looking at establishing a lightweight open banking data store that can maintain external data, and use existing data services to access enterprise data. Alternatively, the model that can potentially be used involves having an open banking data store with enterprise as well as the external data. Such a data store will serve as the layer to reduce any impact to the existing enterprise SOR applications with increased access to customer data.
Key considerations for open banking data fabric
- Provisioning enterprise data and insights for third parties without any impact on SOR performance or availability
- The ability to ingest and store data consumed through open banking APIs
- A metadata-based framework for ingestion and consent-based data management
- Leveraging existing investments in data platforms to support open banking data needs
- Analytical capabilities to provide real-time insights and opportunities for data monetization
- Smart data packages leveraging artificial intelligence or machine-learning capabilities
- Providing deep customer insights, predictive capabilities to prevent frauds and help innovate newer services
- Flexible mechanisms to integrate internal and external data
- Data lineage and governance for ensuring compliance
- Supporting a high degree of data agility and intelligence
The realization of a data fabric can be achieved through a data-centric architecture built on a big data platform. This multi-layered platform can help meet all the above needs organically, while providing support for varied nature of data in terms of speed, type and processing needs. Here’s a round-up of data fabric layers:
- Ingestion layer: to take in enterprise as well as external data through streaming as well as in batches supported by various integration methods. The integration with enterprise information systems will have stream as well as batch exchange capabilities to push and synchronize relevant data onto a data hub.
- Data hub layer: to support storage of all shared customer data across domains (product data, other partner data, market information and digital behavior data from social feeds) – all within this massive distributed storage.
- Analytics layer: for deriving insights from the data hub to support a holistic customer view, drive customer journey. This would also have prediction capabilities to prevent fraud and help product and service innovations.
- Data access layer: to support display of business data views to multi-channel interfaces, including third-party access.
Governance layer: to govern metadata management, data lineage and access controls etc.
Our vision of building such a data fabric will bring significant benefits to the enterprise including a consolidated view of the customer and, insights that could help innovation in new customized products and services for individual needs. In addition, it could provide relief from massive and expensive queries to legacy systems, and support various data formats and sources.