Building a Scalable Fast
Data Platform to Expand Digital
Financial Services

Working on-site with the DBS team to utilize cloud-native technologies and provide customers with an innovative solution

To help DBS tackle the challenges associated with the low efficiency of the banking systems and the lack of consistency in building data models, Grape Up delegated an experienced development team to collaborate on-site with the customer. Teams used agile practices, such as sprints, Kanban, according to the current customer’s needs.

By building a scalable and robust fast data library, Grape Up empowered DBS to utilize various data sources. To achieve a determined goal, the development team build a system consisting of three components:

● Normalizer – here, all the data are consumed and normalized, this element is also responsible for publishing messages to Kafka.
● Updater – this part consumes normalized messages and saves data in the bity format in the Aerospike database (Aerospike was chosen due to its high scalability and quick access to NoSQL databases) to avoid time losses needed for deserialization.
● RPC Server – RPC API for submitting requests to the cache.

The entire stack of components was generic and fully configurable through config-server, environment variables, and provided rules (e.g., normalization). The components building the library were used in applications deployed on Cloud Foundry, which allows easy scaling individual components according to the project needs, such as increasing the number of application instances using RPC Server while there were more domain readings than writings.

Unifying API and data models to provide innovative financial services for customers on various markets

Solutions built by Grape Up and the customer’s team empowers DBS to provide a fundamental change in their services – all its clients have access to the latest, most innovative services. By unifying API and data models, DBS gained the ability to deliver the same customer-facing applications to its customers, regardless of the system they use and their location. Previously, the leading Asian bank found it challenging to provide numerous apps as each region had different deployments based on local regulations and the period when those systems were built.

Like many other financial enterprises, DBS struggled to develop new features and digital products duo to the legacy backends. As these enterprises grow at a rapid pace, both by incorporating various companies with their solutions and by building custom software for new markets. Now, by utilizing the cache layer, DBS can provide the same UI and features developed on top of numerous backend architectures.

Such a solution improves customer experience, accelerate time-to-market for new products and services, and reduces costs of deployments and maintenance.