
Building at the speed of AI
Originally published with our partners, Women in Tech forum UK, on LinkedIn here.
Most engineers, designers and product managers spend their careers improving products that already exist. Far fewer get to build one from nothing.
At Xero, a team in the UK did exactly that. They rebuilt Xero’s UK payroll software from the ground up, new architecture, new design system, new calculation engine and unveiled it at Xerocon London this year as Xero Next Gen Payroll. It is now in beta with real businesses running real pay runs.
Payroll is unforgiving work. It has to be exactly right, every time, for every employee, under rules set by HMRC that leave no room for approximation. Rebuilding it meant getting the hard parts right while reimagining what the experience could feel like and doing it at a pace that would have been difficult to imagine a few years ago, because the team leaned into AI to accelerate how they worked. We asked three of the women at Xero who led the design, engineering and product on that build what it actually took.

The hardest problems weren’t the ones you’d expect
Ask an engineer or a designer about the most complex thing they solved and you might expect a technical answer. The answers here were mostly about people, alignment and assumptions.

A day in the life during launch
A new product launch is not a single moment. It is a stretch of days where everything the team has built meets real customers for the first time, and the work shifts from building to watching, checking and deciding quickly.

What building with AI changed
When it comes to building with AI, the interesting thing is not that the team used AI. It is that adapting to it was itself one of the hardest parts of the job, a question of judgement rather than tooling.
The challenge was working out which long-held practices still serve you and which are simply habits. Natasha frames it as carrying forward hard-won experience without letting old assumptions close off new possibilities. Tigerlily describes improving AI ways of working across the payroll team as a deliberate piece of work in its own right, sitting alongside the design itself.
The discipline runs the other way too. The payroll calculations remain deterministic, because they have to be exactly right. AI accelerated how the team worked; it did not lower the bar for what the product had to get correct.
Advice for women growing their careers in Product and Technology

What gets built next
Xero Next Gen Payroll is one example of the kind of work Xero's UK product and tech teams are getting to do right now. A large-scale, fast-paced build on a product that thousands of small businesses depend on to pay their people. It also sits right at the heart of a bigger shift: AI is fundamentally changing what's possible, both in how we build and what we can build. The question has shifted from whether something can be built, to what is most worth building. This project required getting closer to the customer, deeply understanding their world and then reimagining how tasks get done.
If you're interested in working alongside people like this at Xero, check out our open roles here. Come and help us explore what's possible.



