Interview: Illuminating the black box

Together, Tim Williams and Peter Gale have spent nearly four decades providing digital solutions to local and central government. As authors of the Actian ebook, ‘Accelerate a Digital Transformation in the UK Public Sector’, they specialise in building data governance systems that can facilitate automated interoperability and supporting local government along that journey.

We spoke to them about knowing what you have, knowing what you need and gaining the small wins needed to snowball large scale digital transformation.

Data Governance

“What are the pillars of data governance?” you ask. Tim tells us it’s “Data quality, data protection and data stewardship. It’s about understanding what data you have, who owns it, what the policies are around it and how you access that information securely to make sure that information complies with regulations. The data that you have needs to be consistent, complete and accurate. Governance can be a dirty word. It’s associated with bureaucracy and policing but the opportunities can be enormous”.

Enterprise sales engineer Peter clarifies that issues with data quality such as in merged data classifications for different attributes emerges from poor data governance and is a frequent issue they have noticed in government. “You need to understand what your data is and what it represents and these must be separated in your data governance. A data catalogue is absolutely fundamental for this and it’s one of the first things we recommend.”

What am I missing?

Three of the most common mistakes made by the public sector when it comes to managing data are overestimating the quality of your data systems, not understanding your needs and band-aid solutions to systemic problems.
“We observed a huge gap in the perception of data readiness between senior executives and the people who are actually responsible for looking after that data,” said Peter. It’s perhaps no surprise then that short-term data quality initiatives are destined for futility. Tim added, “People think that if they implement a one-off data quality programme through standardisation or removing duplicates that their data is now ‘clean’. If you haven’t fixed where that dirty data is coming from, then the problem will just come back again. You need to have controls in place to constantly control data quality.”

Crossing wires

Progress has also been made in creating systems for cross-agency integration. “The NHS has been very successful in creating integrated healthcare systems. “Within the NHS Trusts, there is now an edict around integrated care systems, which is not just the central point of healthcare access for specific services like going to A&E, but these trusts now need to work with a multitude of different healthcare providers to deliver better services for patients. That is only enabled by better data sharing and as they’ve digitised they’ve become much better at that,” Tim told us. 

In agreement, Peter added, “Even in this leading case study, they haven’t built robust processes around that data sharing, certainly not to a point where it’s automated. There is risk aversion to cross-agency sharing due to privacy concerns. What happens when it goes wrong? The Police Service of Northern Ireland (PSNI) data breach was a good example of when it goes wrong. Often, the data that departments want to share will be in different formats. You do need technology to help you standardise data but crucially you need the people and the process to make that technology work.” 

Budgets

With all this talk of transformation, we wanted to ask Tim and Peter the pragmatic question that has been raised in our recent webinar: What about budgets? Tim told us, “Local governments are being constantly harangued to deliver more with less money and less help. The public sector is so important not just for keeping the wheels turning in the country but also because people’s lives depend on the public sector and they’re being asked to do things in an increasingly difficult environment. That’s part of the reason why we published our e-book ‘Accelerate a Digital Transformation in the UK Public Sector’”.

Extending on this point about the importance of understanding the problem first, Peter elaborated that, “All transformations have budgets behind them and so you have to understand what the problems are first before you give into a shiny new technology for the sake of it. If you want to save money, you want to find a solution that solves most of your requirements because you’ll never find a solution that does 100%.”

Tim emphasises the value of phased implementation. He told us ‘“you can’t boil the ocean”. Whilst the big vision is necessary, such as a single view of ‘citizen’, “the governance needed to place adequate controls on personally identifiable information and the challenge of bringing together complex information in different formats is a mammoth job. Starting with the use cases that are going to add immediate value whilst still contributing to that long term vision is the best way forward.”

Roadblocks to software integration

We wanted to know what the main barriers to software integration were. “There is a proliferation of legacy across government. There are examples of data management systems that run critical services and have been in place since the 90s and become totally black box. Nobody knows what would happen if you changed it. That’s very difficult to plug into a new cloud technology which isn’t going to play nicely with a very old system.”

On the cloud, Peter told us, “There’s got to be a good reason to go to the cloud. On the one hand, a lot of cloud systems are much more secure. They have multiple levels of security and authentication. The flip side of that is the cost. People in the past have looked at the cloud and thought, great I can get rid of my on-premises data centre. I don't need to own the hardware, but then they lose control of the cost. They don’t manage access to the cloud system properly. Choose the cloud carefully and when it solves a problem, both technically and in terms of the business need.

Game changers?

Everyone is saying that Generative AI is going to be a game changer. Hold your horses is the message from Tim and Peter. Unlocking the benefits of GenAI is totally dependent on data quality. Without the vision and the data management infrastructure to support quality and agility, generative AI will be both out of reach and undesirable.

Read Tim and Peter's Actian ebook, Accelerate a Digital Transformation in the UK Public Sector