Getting data right is not an exciting pursuit, says Dan Klein, global chief of data & AI at Zühlke, the Swiss technology giant. “It’s just hard engineering, but it’s super important,” he enthuses.
Klein, who has worked the world over, consulting on some of the biggest data-driven innovation projects, be it shipping or energy grids, believes Scotland is investing in data in a way that is eclipsing other nations.
But at the same time, he worries government-sponsored programmes supporting societal and mission-orientated goals are bypassing some of the fundamentals.
“We have a tendency to fixate on all the cool artificial intelligence (AI) and science-y stuff,” he adds. “But the real mechanics of data, the messiness, and the things you need to sort out underlies all that. It’s a plumbing issue, essentially.”
I speak on Teams to Klein who now lives in Ardnamurchan. It is perhaps an unlikely choice for someone working at a multinational firm, but with a gigabyte fibre connection and “beautiful” surrounds, his work-life balance seems assured.
Zühlke opened an Edinburgh office in June 2022, with some fanfare around its intentions to break into Scotland’s public sector IT realm.
The company is not a known commodity here, but in Switzerland its precision technology literally keeps the trains running on time – to the second.
It’s that dedication to using data that will win through in Scotland, too, says Klein, where he sees alignment between Zühlke and government data strategies.
Klein says: “It’s not exciting, but that’s where the money needs to be spent if we are going to get frontline public services to work better. For example, in the discussions we’re having with the Scottish Ambulance Service, we see that, as in so many organisations, the information is siloed in five disparate datasets.”
Klein highlights how it’s only in critical situations, like someone dying – that an inquiry ensues – and the data is brought together.
As a routine, he adds, the call handling data, onboard ambulance systems and data from the receiving hospital all exist in different formats and settings with very little, if any, data shared.
So, what it is that prevents public sector organisations from getting their data houses in order?
Klein says: “It’s many things. Overall, government departments, councils, they’re actually quite good at sharing data – when required. And that’s the bit that needs focus.
“The main issue is that data is hard work, and there are skills shortages that stem from that, which makes it really challenging. And then there’s legacy kit that is difficult to work with.”
He adds: “And if you’re a career civil servant, you’re probably not going to make your progression on the back of trying to sort out a data project. It all takes time, money and commitment.”
That’s where Zühlke comes in. The company played a critical role in the design and development of the NHS England Covid-19 app and supporting infrastructure, a project that went from ideation to national launch in 12 weeks.
“With the Covid app, we came in after the Isle of Wight trial had failed. One of our unique selling points is that we operate in highly regulated markets, and we have our own internal certification body.
“And that’s what they needed, so we got the Covid app certified as a Class 1 medical device. We then open sourced everything.”
He adds: “We won’t insist on going in to build a product or extract your data and do something else with it. That’s not what we do; we have access to thousands of consulting engineers who go in and work with the teams in-house to get the data flowing.”
Klein admits that Scotland’s geography creates service design issues.
“Scotland is bifurcated, geographically speaking,” he says. “You’ve got people in the Central Belt and then there’s non-Central Belt. The lives and experiences of people in those regions are immeasurably different, so – and I’m using data language here – there are clearly two cohorts who need different approaches in terms of services.”
He adds: “And I know I’m speaking about health here, but the principle is once you have got the data right, and you move towards data maturity, that’s what enables you to do all the clever stuff, whether it’s machine learning or deep learning.
“I love the fact that we’ve got all this great research at the universities of Glasgow and Edinburgh, and the rest. But they’re about four steps ahead of where the likes of the Scottish Ambulance Service are. We’ll only be able to do machine learning when we get the basics right first.”
Partner Content in association with Zühlke