In September 2026 we took a single question to the Manchester UKHSA Conference, printed it on a wall, and handed out pens. Close to a hundred people stopped by our stand, and what they wrote down turned out to be a remarkably clear brief.
The question was deliberately open: What improvement in your data would enable your biggest ambition?
What the wall was really saying
Ask a room of public-health professionals about their biggest ambition and you might expect the answers to reach for the frontier: predictive models, agentic tools, artificial intelligence. A few did, but for the most part the answers were strikingly consistent, and strikingly practical.
People asked for:
data that is joined up
data they can actually reach
data they can trust
data they can share safely across organisations and borders. And, more than once:
data they can simply find in the first place.
The wall, in other words, wasn’t really about ambition at all. It was about the conditions that ambition depends on.
Even the notes that did reach for AI were data problems in AI’s clothing. Get ahead of outbreaks. Real-time early warning. Target interventions. Insights, not just numbers. These aren’t model problems, they are ambitions that only become possible once the data that is being consumed is connected, current and trusted.
We like to think of ambitions as the rocket. The foundations are the launchpad. On the conference wall, almost everyone was asking for the launchpad; a quiet argument for investing in what sits underneath the shiny models.
Six needs beneath one ambition
Grouped together, the wall read like a specification for a modern public-health data capability, with six needs repeated. Whilst five are foundational, the sixth is the ambition they all serve.
Join it up
The largest and loudest theme by some distance. A single, connected view of data; across systems, across organisations, and across a person's whole health journey. Shared identifiers. One source of truth. An end to the everyday absurdity of two systems that hold related data and can’t speak to each other.
“Want to compare vaccine uptake with other data, but the systems can't talk to each other.”
Make it reachable
Joined-up data is worth little if getting to it is a fight. People wanted the right data quickly and safely, without wrestling the platform that holds it.
“Access to the data I need, when I need it.”
Make it trustworthy
Accuracy, completeness and integrity ran through everything, because insight is only ever as good as the data beneath it. Duplication, gaps and disagreements over definitions were named as the things that slow the journey from data to decision.
“More accurate data collection I can trust.”
Govern it and share it, safely
Sharing across the NHS, local authorities, central government, and beyond that’s done securely, lawfully, and with clear stewardship so that governance isn’t a brake but an accelerator that makes sharing possible.
“Manage it securely and lawfully.”
Know what you have
A quieter theme, but a telling one. You can’t join up, trust or govern data you can’t even see. Several people simply wished for a way to know what already exists, and what that means.
“A data catalogue, so I know what we have.”
Then, act
This is the ambition the other five serve: real-time early warning, targeted intervention, and analytics that lead somewhere rather than stopping at a dashboard. It's the right ambition but has to be grounded in solid foundations. It’s the argument we make in full in our whitepaper, AI is the rocket, foundations are the launchpad.
“Get ahead of outbreaks — trends and warning signals.”
Transform have been here before
None of this is theoretical for us. We've spent years doing exactly this kind of foundational work, even inside of UKHSA.
Our work with UKHSA: Enabling a seamless health data transition
£30m transition programme
32TB sensitive data migrated, IG-compliant
partnering with UKHSA since 2020
Our work with HEE: Turning 10 million records into a model that predicts attrition
10m records unified from multiple systems
60%+ accuracy predicting who's at risk
£250k invested per doctor trained
Where to start?
We wrote a whitepaper around foundations for AI so you may want to take a look at it here. And if one of these foundations is the thing quietly holding your biggest ambition back, we'd value a conversation about it - reach out at transformation@TransformUK.com.