Endowus was founded in 2017 by former global investment bankers who were frustrated at the inefficiency, difficulty, and cost of investing their personal pension monies in Singapore. Detecting an opportunity, they set out to help educate Singaporeans about opportunities to improve their financial outcomes while providing them better tools to direct the investment of their own retirement savings through government-approved programs.

Today, Endowus is Asia’s leading fee-only digital wealth platform, featuring access to superior investment products, personalized advice, and lower costs on a seamless digital investment platform for all investors. Licensed by the Monetary Authority of Singapore (MAS), the company was the first digital advisor to span private investment accounts along with Singapore’s public pension and retirement savings accounts. By the end of 2021, Endowus had grown to more than S$1.5 billion in assets.

The Challenge

From the start, Endowus sought to put customers first and understand them more holistically through data. To lay this foundation, Endowus built a cloud-native platform based on an event sourcing pattern with a microservices architecture (MSA) in which all applications run on containers.

With this architecture, every interaction between the company’s platform and its customers, including the transactions occurring as a result between its internal microservices, is recorded as a form of event in Endowus’ Apache Cassandra® database. Events can be triggered, for example, by a customer clicking a button on the company’s website or mobile app, making a deposit or withdrawal, or selecting an investment within the platform.

“The event sourcing approach that we chose produces tons of data that is always flowing and accumulating, so we had to be very mindful about the workload requirements in terms of database write performance and read performance. Because of this, from our earliest days, we started relying heavily on proven open source technologies such as Cassandra,” says Joo Lee, CTO and Partner, Endowus. The Cassandra database allows Endowus to store exponentially growing volumes of unstructured data, deferring data structuring decisions to more flexibly meet evolving business requirements over time.

Initially, the company hosted its own Cassandra environment, but as Endowus’ business grew, the IT team looked for additional support. “Cassandra scales beautifully, but the infrastructure management burden became a challenge for our team of engineers, who are always busy developing better, data-intensive applications to make personalized investing more accessible, affordable, and convenient,” says Lee.

The Solution

To outsource the management of its Cassandra database environment, Endowus turned to DataStax. Built on Kubernetes, DataStax Astra DB provides Endowus enterprise-grade features for its Cassandra environment, in addition to the efficiency and flexibility of a cloud-agnostic managed database service. The solution offers security for data at rest and in motion, role-based access control, and single sign-on, plus configurable auditing, log scanning, and filtering to help prevent and detect potential breaches. The Astra DB pay-as-you-go pricing model enables Endowus to pay by reads, writes, and storage in order to align costs with the growth of the business.

The solution leverages DataStax contributions to the Stargate Data API Gateway, which eliminates drivers and the need for Endowus’ developers to learn Cassandra Query Language (CQL). With Stargate, the team can use a unified gateway and modern APIs including schemaless JSON, REST, and GraphQL for streamlined development. The Endowus team can also use the new open source Remote Procedure Call (gRPC) for Apache Cassandra that was developed by DataStax. Based on the gRPC framework created by Google, the gRPC API enables a cloud-native design pattern and supports advanced features of HTTP/2 for improved performance.

The Results

Lee is gratified that his team can now spend more time developing customer-centric services and less time worrying about supporting infrastructure. “Since we’ve expanded quickly, we appreciate being able to rely on DataStax and Astra DB to alleviate the administrative burden of our database,” says Lee. “Also, we value the resiliency of DataStax and Cassandra. We have to ensure very high availability for our customers, as mandated by our regulator,” he says.

Endowus is realizing the benefits of faster data to enable its developers to provide better investment customer experiences.

Data insights fuel explanatory analysis that Endowus values to understand the hows and the whys of various milestones in its customers’ journeys, from exploration of its investment services to onboarding and engagement over time. “We can closely track various customer groups based on descriptive data to quickly examine impacts of particular programs or investments. For example, we used explanatory analysis to find out at which stage of the onboarding process our customers were most likely to drop or get stuck. We could then focus our efforts on improving that part of the customer journey,” Lee says.

“It is extremely important that we minimize the time to data-driven business decisions by accelerating data collection and analysis that generate actionable operational insights. This is what gives us a competitive advantage in this market. Faster data is critical to that,” he says.

“With the high performance of Astra DB, we can rapidly mine our voluminous data to deliver tremendous insights about our customers to business managers across our organization, which helps further propel our mission and our growth,” says Lee.