Architecting for agility and responsiveness

Technology changes are impacting the fabric of society—from shifts to a more fluid workforce to new agile business models that change government operations as we know them.

Public service is evolving every day. Agencies are tasked with managing new policy changes, everchanging citizen demands and technology advancements that pave the way to better service delivery.

Responding to these changes is challenging given that some organisations are using decades-old systems. Applications are often strung together like spaghetti and making one simple change requires the same change to be made across numerous systems. It is too risky for government to rip and replace these core systems to modernise for the future. But it is even riskier to sit idle and not evolve systems to keep pace with a fast-changing world.

There is a way to make change less risky, less costly and incremental. Creating an IT architecture with greater agility enables public service organisations to make their technology landscape more fluid and flexible, setting them up to be more responsive and adaptable to future changes. It also reduces the risk of large modernisation techniques while delivering the benefits of a less complex infrastructure. The good news is that this approach builds on the core of what the organisation has already established. And no matter where an agency is on its technology journey, there are multiple ways to begin the transformation.

The six shifts outlined below show how change can happen incrementally to keep core systems intact while leading public service toward a modern architecture. Following these principles can reduce the amount of risk while managing future change, and it allows the organisation to rapidly adopt emerging technology at the pace of change.

Digital decoupling

The digital landscape will continue to change rapidly, so public service must adapt to new demands and adopt new approaches just as quickly.

Digital has affected the way citizens engage with services throughout their lives, so it drives higher expectations for interactions with public service organisations. Agencies must meet customer demands—at speed—but current IT systems are tightly coupled, making it difficult to quickly modify the customer experience. Agencies should implement decoupling strategies to unlock legacy IT's value. For example, in the US, Accenture Federal Services’ Digital Decoupling Survey identified how to break down the barriers to innovation and new technologies; finding that 87% of IT leaders say they want the best of new IT, but only 36% have a clear strategy for doing so.1

Digital decoupling separates the user experience layer from core IT processing to enable greater flexibility and simplified user experiences. Using technology enablers, such as application programming interface (API)-based integration and robotic process automation, public services organisations can offer new digital services without fundamentally changing core systems. For instance, decoupling the core processing system from a postal agency’s citizen-facing website enables the organisation to work with public or private partners to create better and more integrated customer delivery experiences.

By decoupling new systems from legacy, the possibilities for rapid and flexible reaction to new technologies and customer demand become endless. Imagine creating a conversational user interface and machine learning to “guide” citizens through visa applications; or offering a personalised mobile web application to give residents individual and real-time access to city services. Such applications that personalise the user experience and radically change the status quo are possible today, but too often held back in implementation due to integration challenges with existing systems, data and processes.

Platform-based architecture

A platform architecture delivers the agility needed to adapt to business requirements without needing to evolve the entire system.

Public service architectures have historically been monolithic application structures that are tightly coupled and therefore difficult to upgrade and modify at scale. Modular, decoupled platforms will allow an organisation to adopt modern capabilities, such as cloud, to improve scalability, flexibility and availability.

Techniques such as domain-driven design can be employed to decompose functions into domains and modules that can be modernised discreetly.

Traditionally related core processing capabilities such as customer and payment systems would be delivered by a single platform, however shifting to a focus on agility – the payments engine and the customer engine may be considered as separate platforms that can have independent architectures, technologies, and data. One may be hosted on premise, the other in the cloud. This approach allows the two to evolve independently and react quickly to their specific needs without impacting the other. APIs and events will allow other platforms to interact with and share data about payments.

Business and IT must align around the platform with both sides serving as platform owners empowered to evolve the architecture. A platform-based architecture allows the team to talk in a combined language to collaborate and change at speed to better serve customers.

Progressive modernisation

Removing constraints at the core frees an organisation to modernise progressively and use analytics to realise value where it counts the most.

Government is rife with monolithic legacy systems that constrain the core and hinder the ability to make discrete changes continuously. Wholesale system replacements are too risky and a large-scale rip and replace just swaps one elephant for another.

A modernised core architecture breaks down the system into logical and discrete parts that can evolve independently. In this way, agencies avoid the risk of inaction and minimise the IT debt that will accrue if a system takes too long to implement.

To effectively break a core legacy system into logical parts, agencies can leverage microservice decoupling for modular functions that can easily be decoupled or in-place modernisation, which is more suited for functions that are less modular and cannot be easily decoupled. Once this is achieved, agencies are able to replace logical parts over time rather than replacing the entire system. This approach also enables adoption of software as a service or other cloud- based services for discrete functions.

If the agency spends time determining the right approach to modernisation, it will do it right the first time and continue to reap the benefits.

As an example, a child safety agency is looking to decompose a custom-built legacy system into independent, loosely coupled functional groups. The system is complex and supports the majority of front-line services and is therefore difficult to modernise. Using this approach, the agency is able to replace each functional group independently, reducing the risk and complexity of the modernisation.

Event-driven architecture

Moving from process-based to event-based orchestration allows for real-time processing.

In many current architectures, services and functions are tightly coupled together as part of an end-to-end business process. Event-driven architectures began with the likes of Spotify and Netflix when the private sector recognised the power of the ability to fluidly react to individual events, such as consumer choices, rather than sticking with a rigid process.

Event-driven architecture decouples functions, allowing them to operate independently. This approach further enables agility, scalability, flexibility and resilience in the architecture.

Modern data and integration technologies, such as data streaming, facilitate ingesting and managing massive amounts of external data, such as payment information, to make processing easier. Event-driven architecture allows an agency to monitor data streams, applications, processes and events with the granularity needed for efficient processing. A system can be configured to listen for certain “events” that affect them and trigger functions when those events occur. This independence allows services to run in parallel and when coupled with modern cloud architectures and DevOps, this parallelism can work at a scale previously not possible.

Applied Intelligence

Unlock value from data in real time based on central modeling.

Applied intelligence combines artificial intelligence with data, analytics and automation under a bold strategic vision to transform public service across functions and processes—at scale. It also shifts the focus from creating insight to action on data and making improvements based on the insight.

Public service agencies can be more proactive by processing massive amounts of data in real time and responding in targeted and meaningful ways. For instance, sophisticated artificial intelligence analytical models allow more intuitive experiences for customers, and machine learning can inform what would be the next-best action to get to better outcomes. In Singapore, AI is helping the Central Narcotics Bureau to catalog and identify evidence from a crime scene, comparing that data to the agency’s database to help make connections that can strengthen cases and fight crime.2

Applying intelligence calls for extending existing data stores to include data lake and analytics capabilities that drive smarter business decisions. Data scientists can further harvest analytical data to uncover insights that were previously unknown.

Agile organisation

An agile organisation—across business and IT—will fully unleash the benefits of architecting for agility.

An agile organisation is the underpinning of agile techniques, technologies and architectures. It is founded on a culture of continuous delivery, innovation and collaboration. Shifting to an agile organisation will be more difficult for some organisations than others because it requires a full-scale move to a new platform along with new governance, skills and processes.

The structure of an agile organisation is centered around platforms and services with clear business ownership and joint business and IT teams that work based on agile approaches and fully leverage DevOps. Team will take on new shapes—such as squads, pods and tribes—to work together fluidly and with agility to achieve an outcome.

Governance will need to change so that these new decentralised teams can work independently, but with strong central control. Teams will require new skills such as innovation management, agile and DevOps. Upfront design and planning will help to avoid unnecessary bottlenecks in rollout or risks due to immature capabilities.

In Australia, the Tax Office is using agile project management techniques across IT and the business, to deliver personalised digital services.3

Enabling agile transformation

Architecting for agility is a fundamental shift in delivering IT services, moving away from operational efficiencies to flexible, decoupled solutions. It can position a public service agency to adapt to future technology changes and everchanging customer needs. These four enablers support the transition, making it easier to get the most value from a modernised technology approach.

The time for change has come. Transforming now will position agencies for agility and stability as they prepare for the future.


API technology allows an organisation to extend its digital edge to include the broader ecosystem of partners. It allows critical shifts such as digital decoupling and platform-based architecture.

Operational automation

DevOps must be in place to enable the core elements of agile architecture. The tools allow an organisation to be agile and rapidly respond to change. Secure DevOps enables agencies to embed security principles and controls into all aspects of the application lifecycle.


Cloud technologies in both private and public models offer new possibilities for architecting and hosting solutions for resilience. Agencies may consider a blend of on-premise private cloud for core processing and public cloud for digital experiences.


The threat vector has increased exponentially due to new channels. To make quick changes to systems to protect the organisation, it is essential to centralise internal identity and access management to reduce operational effort.

View All

David Thatcher

Lead – Technology, Consulting, Health and Public Service, Australia and New Zealand

Peter Thomson



Public service agility: Unleashing innovation
Championing the NEW public service tech leader

Subscription Center
Stay in the know with our newsletter Stay in the know with our newsletter