Agility in the design and development of CBDCs: From discovery through launch

Many central banks globally are researching the possibility of launching a central bank digital currency (CBDC). Amazon Web Services (AWS) has worked with customers in the financial industry as they design secure, resilient, and scalable critical financial infrastructure networks. From this experience, AWS has developed best practices from the field of modern application development, well-architected principles, and critical infrastructure development to support the creation of a fit-for-purpose CBDC.

These best practices are collected in a recently released whitepaper, the third in the AWS CBDC whitepaper series. Part 3: CBDC: Agility in design explores and explains how to bring these best practices together through an agile development lifecycle for a CBDC. This framework aligns the design of a CBDC to the goals of a particular central bank and its stakeholders. Additionally, each stage throughout this process encourages reflection on a central bank’s policy vision and only recommends proceeding with a CBDC if alignment remains with all relevant policy goals.

In this post, we share key best practices for central banks to consider as they move through the CBDC development lifecycle (Figure 1). For deeper technical insight and considerations for each stage of the CBDC development lifecycle, read the full whitepaper.

Figure 1. The agile CBDC development lifecycle, in which Stage 1 is the discovery and prioritization phase, Stage 2 is for experimentation and design, Stage 3 is the development of a proof-of-concept, Stage 4 features the launch of phased pilots, and Stage 5 is the launch of incremental improvements.

1. Embracing experimentation in CBDCs

Working backwards from a clear vision is paramount. From this vision, central banks can develop clear objectives that drive the evaluation of a CBDC project and shape its development. These objectives vary substantially across the wide spectrum of countries due to variation across levels of digital infrastructure, financial system maturity, economic goals, and more. Thus, experimentation is needed to make sure CBDC design will meet each set of unique requirements. Additionally, at each step of the journey, central banks can return to their vision, the related CBDC objectives, and assess whether the system as designed meets that vision.

Hypotheses development and experimentation can be completed in many ways. Some central banks have encouraged and rewarded private sector research by hosting public CBDC challenges aimed at finding technical solutions to key obstacles. These include the MAS Global CBDC Challenge and Brazil’s LIFT Challenge Digital Real Edition. Others have leveraged partnerships with academic institutions, such as the MIT Digital Currency Initiative, while others focus on internal research that uses design frameworks, such as that proposed in the Oliver Wyman Forum and AWS whitepaper, Retail Central Bank Digital Currency: From Vision to Design. These tools help gather initial insight and strategically selects areas of further exploration.

2. Designing for modularity and optimization

Multiple central banks are currently in the proof-of-concept (PoC) stage of their CBDC research. PoCs should leverage best practices to test system efficacy and design. This is done by building and testing the fundamental pieces of the CBDC system. To do this, we recommend three development best practices:

1. Design a modular architecture, which separates core functionality into independent and interchangeable units. This allows both for the testing of components individually, but also the swapping out of components to test for alternate solutions.
2. Design with automated scalability and fault tolerance in mind to decrease dependence on manual system intervention. Consider event-driven architectures.
3. Design datastores to be decoupled from computation and optimized for use to limit the impact of operational failure as well as increase system efficiency.

3. Building for security, resiliency, and scalability

With a CBDC architecture aligned to desired function, three key technical requirements of scalability, resiliency, and security must be tested and optimized. Strong security provides protection from cyberattacks and unauthorized use. AWS services are designed to be secure, and AWS provides a suite of security tools for customers to handle access, authentication, incident response, and more—as well as security services that help support meeting various regulatory compliance programs.

Automated scalability addresses both the need to process a high number of transactions and supports them being processed in near real-time. A design with autoscaling allows each functional component to grow or shrink with demand. Additionally, this can create a more cost-effective system, which reduces per person servicing costs.

High resiliency includes high availability with the 24/7/365 ability to make payments, plus the automatic recovery from operational failure, infrastructure disruptions, and natural disasters. The AWS Cloud is architected to be highly scalable as well as resilient. If disaster were to strike one part of the infrastructure, services are built to recover with minimal manual intervention.

4. Anticipating continual innovation

While agile CBDC development is depicted as a linear process, optimization of the system requires a continual process with new learnings, leading to new ideas, which may launch the development process again to incorporate new functionality. This necessitates designing for incremental deployments, which can be supported through a microservice architecture and continuous integration and continuous delivery (CI/CD) processes.

 Microservices ease the building and testing of new functionality that can be released separately and even in parallel. This supports increased experimentation and a reduction in deployment-related risks. And leveraging CI/CD principles can accelerate delivery, reduce deployment-related downtime, and provide additional insights from increased monitoring. Additionally, AWS provides CI/CD tools that automate the deployment pipeline, such as AWS CodePipeline, which can improve consistency and ease of management of code deployments.

Beyond designing for continual innovation, potential use cases can also be explored that support functionality that is beneficial to both system operators and end-users. Two use cases that support these customer sets, respectively, are advanced data analytics and internet-of-things (IoT)-generated micropayments.

The power of data and advanced analytical tools is immense. Insights derived from aggregate CBDC transaction data can support the monitoring of economic conditions in near real time, allowing central banks to be proactive instead of reactive in their monetary policy actions. The development of sophisticated models to identify criminal activities can be built with artificial intelligence (AI) and machine learning (ML) tools to identify previously undetected malicious trends or behaviors. To take the immense amount of data produced and turn it into actionable insights, we propose a data analytics architecture described fully in Appendix D of the whitepaper.

Figure 2. An analytics data flow architecture, explained in more detail in Appendix D of the whitepaper.

Though IoT-enabled devices are ushering in an era of machine-to-machine payments, the ability to cost-effectively process these payments is still lacking. A CBDC system designed to efficiently process IoT-generated micropayments can fill that gap for use cases ranging from utilities to media purchases and more. Leveraging AWS IoT Greengrass, an open-source edge runtime and cloud service for building, deploying, and managing device software, we present and explain an example IoT architecture in Appendix E of the whitepaper.

Figure 3. An IoT micropayment transaction flow architecture example, explained in detail in Appendix E of the whitepaper.

The creation of value-add use cases such as those enabled by analytics, IoT, and other innovative technologies can create the product-market fit needed to generate the desired usage amongst target user segments.

Looking forward for CBDCs

Preparing for a future world with multiple CBDCs in deployment is a journey that most central banks have already begun. AWS can help you design services that add value to a CBDC system by offering the domain expertise needed to help you explore and implement them. AWS provides multiple offerings for each stage of CBDC development, from customer advisory sessions to support the discovery and prioritization phase of Stage 1, to Working Backwards sessions in the experiment and design phase of Stage 2, architecture development support in Stages 3 and 4, and more.

Figure 4. The CBDC agile development lifecycle with AWS offerings.

For more information, browse AWS technical resources to expand your knowledge of the AWS Cloud or contact us to connect with an AWS expert.

Read more about AWS for financial services:

• Approaches in retail central bank digital currency: Aligning technology with policy
• The future of money is digital: How the cloud can deliver solutions for central bank digital currencies
• Emerging economies will lead the charge for central bank digital currencies, say experts
• How the cloud can enable enhanced risk management and innovation

Subscribe to the AWS Public Sector Blog newsletter to get the latest in AWS tools, solutions, and innovations from the public sector delivered to your inbox, or contact us.

Please take a few minutes to share insights regarding your experience with the AWS Public Sector Blog in this survey, and we’ll use feedback from the survey to create more content aligned with the preferences of our readers.