Government agencies around the world are facing an expectation and implementation gap in joining together and acting upon the data that they hold. Like in the private sector, government agencies need to digitally transform, build resilience and modernise legacy technologies. Doing so is critical to helping them to make better decisions that facilitate better outcomes for citizens and those who interact with the public sector. Just as for private sector businesses, the biggest obstacle to making informed decisions in government is fragmented and siloed data.
A common thread between different government agencies is the need to make decisions at speed. Leaders are often making critical decisions about securing borders, preventing terror threats and dealing with the multitude of problems agencies face every day. A further data challenge for government agencies is that they need to integrate with the private sector.
Whether services are outsourced or privatised, agencies and organizations themselves cannot exist in a silo, leading to an extra level of data unification challenges.
Building a strong foundation for decision intelligence
Gartner predicts that by 2024, 60 percent of government AI and data analytics investment will directly impact real-time operational decisions and outcomes. Decision Intelligence (DI) is the method of delivering on these objectives. DI helps organisations transform their data – which may contain mistakes, inconsistencies or be deliberately misleading – into accurate and complete data that can be trusted.
With a strong data foundation, agencies can build context through networked relationships by deploying machine learning (ML) and artificial intelligence (AI) to automate and improve decision-making. In Gartner’s CIO and Technology Executive Survey 2022, 59 percent of government CIOs say the AI and ML capabilities were already or would be deployed within 24 months. Whilst these technologies are vital to interpreting data, efforts will be seriously hampered without cleaning and resolving quality issues in the contextual data that underpins it.
Decision intelligence in practice
The British government’s bounce back scheme provided an important lifeline to businesses impacted by lockdown. But as much as £4.9 billion of the £47 billion lent by banks to 1.1 million companies has been lost to fraudsters, according to the government’s estimates. In response, the Cabinet Office adopted entity resolution network analytics to detect and fight this type of crime. Fraudulent claims ranged from the sophisticated to relatively simple deception methods, such as applying for different loans with a slightly different address on the forms. But even simple deception is difficult to spot by human agents when we consider that nearly 1.56 million businesses were approved for finance with the Bounce Back Loan Scheme.
Decision intelligence frees civil servants from spending time and effort on manually checking data for inaccuracies or deliberate deception. It can then spot patterns in applications, like numerous applications from similar but different addresses, that would have been missed or taken humans hours to spot. With 1.56 applications to check, without AI and ML it would be a near impossible task.
Where public sector agencies can put DI to work
The NHS database holds the medical records of 65 million people, which along with the unique 10-digit code attached to every patient, means that the health service theoretically has a complete picture of a patient’s history. However, theory currently does not make practice. Swathes of the NHS is still paper-based. And, departments often do not share notes with each other. Moreover, researchers engaged in clinical studies and trials cannot link the data they collect with their patients’ health records.
Decision intelligence technology will stitch together the billions of fragmented data points inside and outside of the NHS. What’s crucial for when private and public sector organizations collaborate is that solutions are built upon an open architecture. This allows organizations to migrate legacy applications and data versus investing in costly upgrades. This approach to decision intelligence will also help ensure that any platform remains in the ownership of the NHS itself, now and in the future.
Managing public expectation
In today’s data-driven world, the public expect the government to keep pace with and implement the latest technologies. This ranges from aspects that directly impact their day-to-day lives such as providing a joined-up record of medical histories to ensuring that taxpayer money does not end up being paid to fraudsters. But government agencies need to manage expectations about what is realistic and within certain time frames. Being able to unify data, address risk and run agencies more efficiently is paramount. But this must be done right, to do right by those who will live by or with its services and those who engage with it in their work.