Discovery Bank enhances client financial health with cutting-edge Azure data and AI technology

As a fully digital financial institution, Discovery Bank is a full-service bank—offering clients access to transactional, lending (including home loans), savings, investment, and forex products. Launched in 2019, it is part of Discovery Group.

Discovery Group, founded in 1992 in South Africa, operates in various sectors including healthcare, life and short-term insurance, long-term savings, wellness, and banking. The company's core purpose of making people healthier and enhancing and protecting their lives forms the foundation of its shared-value model, which creates economic value by addressing societal needs and challenges. For example, in health and life insurance, the group focuses on making people physically and mentally healthier, reducing claims, generating rewards for clients, and as such benefiting society. Discovery Bank clients are encouraged to adopt behaviors that help them manage money well, like saving and having adequate insurance. Discovery impacts over 42 million customers globally, supported by more than 13,800 employees across 41 markets. 

Data as a cornerstone of shared value

"Our bank was built on our group’s philosophy of shared value. This means helping clients improve their financial health is core to our focus, which in turn reduces defaults, increases savings rates, and creates profitability for the bank,” explains Stuart Emslie, Head of Actuarial and Data Science at Discovery Bank. “As our clients improve their financial health, this unlocks profitability for us as a bank which in turn allows us to incentivize healthy financial behaviors through rewards paid to our clients. This creates a virtuous cycle and benefits society as a whole."

From the beginning, Discovery Bank has embraced a data-driven approach, ensuring analytics inform every aspect of its operations. This strategy enables the bank to provide personalized experiences for its customers. "Everyone’s financial health journey is unique. Meeting our clients’ needs requires a hyper-personalized approach, not a one-size-fits-all strategy," says Emslie. "We needed best-in-class technology to achieve this."

Choosing the right data platform

When Discovery Bank evaluated the various technology options, Microsoft stood out. “We were particularly drawn to Microsoft’s philosophy of enablement. Their approach is very much about empowering customers to build tailored solutions,” Emslie explains. "Their strategic focus on artificial intelligence (AI) and machine learning (ML) innovation were also key reasons we chose Microsoft.”

Microsoft solutions empower Discovery Bank to find the balance between flexibility and simplicity. “These qualities are often seen as contradictory because flexibility can introduce complexity. However, Microsoft Azure provides a platform that allows us to design and build exactly what we need without creating a significant maintenance burden,” shares Emslie. High speed was equally important. “We wanted a platform that would allow us to build quickly because we have ambitious targets as an organization,” adds Adit Mehta, Head of Machine Learning Operations at Discovery Bank. “For us, the optimal technology stack was a combination of Microsoft Azure and Azure Databricks. Azure’s comprehensive service offering, from infrastructure to AI services, allowed us to craft a robust data and AI architecture at high speed.”

Building a tailored technology stack

Discovery Bank started building its technology stack to address evolving needs and challenges. It began with migrating data to Azure Databricks, which became the foundation for analytics and data engineering. This naturally led to the adoption of Azure Data Factory, a tool that simplifies extract, transform, and load (ETL) processes for moving and preparing data.

As the bank’s workflows matured, the team introduced Azure DevOps to improve software engineering practices like version control and automated deployment. For custom applications with tailored front ends, Azure App Services provided a seamless way to build and integrate these solutions.

To handle growth, the bank adopted a microservices architecture, breaking applications into smaller, independent components. This made it easier to change specific functions without disrupting others. Azure Container Apps facilitated the management of these components, while Azure API Management ensured secure and efficient communication between services.

Most recently, the bank incorporated Azure Event Hubs for real-time event streaming, enabling quick processing of large data volumes and responsive operations. Together, these tools create a scalable, flexible, and secure technology environment tailored to the bank’s needs.

Azure OpenAI was a critical moment, as described by Emslie. “By the time generative AI services were released, we knew they were going to be a game changer. When we discovered how Azure OpenAI provides a layer of confidentiality and the ability to use this technology securely, we realized this would be a huge turning point for creating the data-driven shared value we envisioned.”

Enhancing client interactions with AI

Using Azure OpenAI Service, Discovery Bank introduced an AI assistant that supports customer service agents. “Before, Discovery Bank agents were not able to access client and product information easily,” explains Mehta. “Now, the assistant provides personalized insights about clients, as well as product and process information—all via a simple chat interface. This helps Discovery Bank agents proactively suggest tailored actions to support the client’s financial goals.”

The AI assistance has also reduced these agents’ training time and enabled them to become knowledgeable more quickly about new products or features. “We have close to 200 Discovery Bank agents, and they handle about 3,000 calls a day,” adds Mehta. “We’ve seen consistently positive feedback from Discovery Bank agents, and over time, the feedback ratings have improved, which means our bankers are finding the Azure OpenAI-powered assistant increasingly valuable.”

Discovery Bank also uses Microsoft solutions to optimize its communications with customers. The bank analyzes vast amounts of customer data to understand customer preferences, behaviors, and financial goals. Azure Databricks is then employed to build predictive models that identify the most effective channels, messages, and timing for engagement. “This ensures that our customers are matched with the right products and solutions to meet their needs,” Emslie says. 

All of this has inevitably led to improved customer satisfaction: “We’ve seen rises in customer satisfaction, retention, and engagement metrics,” adds Emslie.

Directly impacting shared value

Discovery Bank measures the impact of its data-driven and AI-powered approach through the shared-value metric. “The shared-value metric looks at two things: the improvement in our client’s financial health and the associated business value,” explains Emslie. “We measure the change in our shared-value metric for each action by comparing it to other actions and a control group to understand the value of each action.”

The bank then converts the shared-value metric into a monetary figure and compares it to the costs of implementing its AI initiatives, calculating return on investment (ROI).  “With the Databricks platform and the Azure OpenAI-powered assistant, we’ve seen a 500% ROI,” Emslie shares. In other words, “Our efforts have created value for both our clients and the business, which aligns with our philosophy of prioritizing shared value over pure profitability.”

Expanding shared value impact

For Mehta, choosing the right partner has proved crucial in this successful journey. "Ultimately, we’ve accomplished our key objective thanks to the technology choices we’ve made: improving client financial health leads to better profitability for the bank, which in turn drives broader client and societal benefits. This is truly a first-of-its-kind banking model.” The support from the partnership has also played a key role. “The relationship we have with the Microsoft team has been invaluable. Whenever we need assistance—whether it’s a subject matter expert, architectural advice, or help with a new service—they are always there to support us. This partnership has been a key enabler for our success.”

As the bank looks ahead, it plans to continue scaling its personalized financial health solutions. “For our next steps, we’re focusing on taking the learnings from what we’ve done on the servicing side and applying them to a customer-facing initiative—specifically, a generative AI-driven assistant for clients, fully built on Azure and Azure Databricks Services. We think it’s going to be a game-changer,” concludes Emslie.

Discover more about Discovery Bank on Facebook, Instagram, LinkedIn, X/Twitter, and YouTube.