Put data behind every decision

Butterfly Data's principal IT consultant, Henry Newman-Burke, and data scientist, Maja Strawinska will be presenting at the Public Sector Connect Data and AI Summit 2026, sharing expert insight on governing data migration across large-scale public sector programmes.

No surprises on moving day: Governing data migration at scale

In their session, “No surprises on moving day: Governing data migration at scale,” Henry and Maja use the familiar analogy of moving house to explore how organisations can successfully manage complex data migrations - particularly when upgrading to smarter, AI-enabled platforms.

Just like a house move, successful migration requires:

• Clear planning and decision-making
• Knowing what data to keep and what to retire
• Accurate labelling through strong metadata and classification
• Solid foundations for future AI capabilities
• Robust governance to prevent disruption

For public sector organisations modernising legacy systems, the risks of poor migration are significant, from compliance issues to service disruption. Strong governance ensures data is not only moved safely, but structured and prepared to power the next generation of digital services.

Supporting AI-ready public services

Together, Henry and Maja will highlight how governance-led data migration reduces risk, protects compliance and enables trustworthy AI adoption across government.

The Public Sector Connect Data and AI Summit 2026 brings together leaders shaping the future of public services. Butterfly Data’s session will provide practical insight for those planning cloud transitions, system upgrades or AI implementation to ensure there are truly no surprises on moving day.

Butterfly Data will be speaking at Digital Leaders AI Public Sector Week 2026, sharing practical guidance on how public sector organisations can strengthen data provenance, lineage and governance to enable trusted AI.

In our session, “Trusted Data for Trusted AI: Getting Data Provenance Right in the Public Sector,” data scientist Maja Strawinska will explore why trustworthy artificial intelligence depends on trustworthy data foundations. As AI adoption accelerates across government, establishing clear data provenance and lineage is becoming essential to transparency, accountability and regulatory assurance.

Why data provenance matters for trusted AI in the public sector

Public sector AI initiatives rely on complex data ecosystems spanning departments, legacy platforms and external partners. Without robust data provenance, understanding where data originates, how it has been transformed and how it is governed, organisations face increased risks around bias, explainability, compliance and auditability.

Strengthening data governance and provenance for AI is therefore not simply a technical task. It is a strategic requirement for responsible AI adoption in government and public services.

What we will share during the AI Public Sector Week

Drawing on experience supporting public sector data projects, Maja will discuss:

• The role of data provenance in trusted and explainable AI
• Common data lineage and governance gaps in government
• Practical approaches to building AI-ready data foundations
• How provenance supports AI assurance, audit and compliance

Join our session

Digital Leaders AI Public Sector Week brings together leaders from government, technology and policy to explore how AI can improve public services. Butterfly Data is proud to contribute insights on building trusted data foundations for AI at scale.

Session: Trusted Data for Trusted AI: Getting Data Provenance Right in the Public Sector
Event: Digital Leaders AI Public Sector Week 2026
Date: 20 March 2026 1:30-1:50PM
Register for free here.

Across the public sector, the pressure to support data-driven decision making has never been greater. Whether you are in a local authority, an NHS body, or a central government department, the expectation is clear: use your data, demonstrate your impact, and make better decisions faster.

So when teams start investing in performance dashboards, it's understandable that the conversation quickly turns to software. Which platform should we use? Tableau, Power BI or another? Which is better for the government?

These are reasonable questions. But they are often the wrong place to start.

The dashboard isn't the strategy

Imagine this: a public sector team has just invested in a new visualisation platform. Licences are purchased, training is booked, and a handful of dashboards are built. Six months later, those dashboards are barely used. Senior leaders still ask for data by email. Decisions are still made on instinct and spreadsheets. And when reports do get read, they are often pages of explanatory text rather than the data itself - making it hard to spot what actually matters or track how things are changing over time.

Sound familiar? It happens more often than it should and the selected platform is rarely the problem.

Building performance dashboards that genuinely support data-driven decision making in government or the public sector requires something the software can't provide on its own: analysts who understand the audience, the context and the question being asked before they open the tool.

What good looks like - regardless of dashboard or visualisation platform

A skilled data analyst working in a public sector environment doesn't begin by choosing a chart type. They begin by asking: who is this dashboard for and what decision does it need to support?

This is something the Ministry of Housing, Communities and Local Government (MHCLG) took seriously when building their own internal performance dashboard. Before writing a line of code or choosing a chart type, their team attended delivery management meetings to observe how existing reports were actually being used and what senior leaders genuinely needed from them. What they found was instructive: senior civil servants weren't looking for reassurance about things on track. They primarily wanted to know where their intervention might be required. Everything else was noise.

A frontline service manager reviewing capacity needs something different from a director preparing a board report. A policy team tracking outcomes needs a different view from a finance team monitoring spend. Getting that right by understanding the audience and designing around their needs is what makes a performance dashboard genuinely useful rather than technically impressive.

The best analysts bring this thinking to every build. They invest time understanding the data model - knowing not just what fields exist, but how data is entered, how it relates to other data, and which of it actually matters to the people reading the report. They translate complex operational data into clear, honest, actionable views. They resist the temptation to include every available metric. They write labels that make sense to a non-specialist at 8am on a Monday. And they iterate with stakeholders until the dashboard earns regular use.

None of that lives in the software.

Where Tableau and Power BI really differ

That said, platform choice does matter, but more so practically, if not fundamentally. For public sector organisations, the decision usually comes down to environment and scale.

For public sector organisations already running Microsoft 365 infrastructure, which covers much of central and local government and the likes of the NHS, Power BI's native integration can offer a practical advantage. Dashboards embed into Teams and SharePoint without additional configuration and access management aligns with existing identity frameworks. Tableau can achieve similar integration, but may require additional display licences to scale that access broadly across an organisation, which is worth factoring in when thinking about how widely you want performance dashboards to be used.

Tableau tends to suit teams doing more investigative, exploratory analytical work. Its visual flexibility and fluid interface reward experienced analysts who need to interrogate complex data before presenting findings. For specialist analytics functions, it remains a highly capable tool.

Some public sector organisations find value in using both: Power BI for operational and organisation-wide performance reporting for public services, Tableau for deeper analytical work. That's not a compromise, it's a sensible reflection of the different needs that exist within a large, complex organisation.

The real enabler of data-driven decision making in government

The most valuable conversation in public sector data teams isn't really about which platform to use. It is about how to build and sustain the analytical capability to make any platform work to its full potential.

Building a culture of data-driven decision making requires more than good software. It requires analysts who understand the organisation's priorities and can connect data to them meaningfully. It requires leaders who trust the dashboards they're shown and know how to act on them. It requires data foundations - quality, governance, consistency - that make the numbers in those dashboards reliable in the first place.

When those things are in place, performance dashboards become genuinely powerful. And the good news is that many public sector organisations already have strong foundations to build on - the opportunity is in making the most of them.

Building dashboards that actually get used

For public sector teams looking to improve the impact of their performance or public service dashboards, the most valuable investment is rarely a new tool. It is developing analysts who ask better questions, understand their audience, and design with the end user in mind - whether that end user is a ward manager, a senior civil servant, or an elected member.

The platform is the medium. The analyst is the craft. And data-driven decision making in government depends far more on the latter than most platform conversations give it credit for.

Butterfly Data works with public sector organisations to build analytical capability, improve data foundations, and design performance dashboards that people actually use. Get in touch to find out how we can help.