Dr Peter Appleby is Head of Data Science and Analytics at Auto Trader, where he brings together deep academic expertise and commercial leadership in AI and advanced analytics.
He began his career in theoretical physics, completing a MPhys at the University of Manchester. Intending to pursue research, but looking beyond physics, he shifted focus to applied mathematics and neural networks, completing a PhD in computational neuroscience at the University of Southampton. His academic career spanned over a decade, with research roles across Germany and the UK working on computational models of neural networks, at a time when the field attracted little of the mainstream attention it now commands.
During this period, Peter was particularly interested in the links between neural networks and the brain, approaching the subject from a computational neuroscience perspective. His work placed him close to many of the formative discussions in the field, many years before the resurgence of large-scale neural networks and the emergence of architectures such as transformers, the technology that underpins large language models like ChatGPT.
Peter then transitioned into industry, taking on commercial data science leadership roles and, more recently, broader AI leadership positions as the technology moved to the centre of organisational strategy. At Autotrader, he operates in an environment that combines the scale and resources of a publicly listed company with the pace and agility of a mid-sized technology organisation.
Peter’s career has effectively brought together two complementary halves: rigorous technical research and applied commercial delivery. That dual perspective shapes his contributions to industry discussions, where he offers a grounded view of AI informed equally by academic depth and real-world execution.
As a data and AI leader, which traits and skills do you think matter most, and which of those have been most influential for you in your current position?
For Peter, effective data and AI leadership rests on a balance between deep technical credibility and the ability to influence decisions in fast-moving conditions. He is clear that leaders cannot step away from the technology. “You need to understand the technology,” he said, adding that his scientific background has been a real advantage. With “tens or even hundreds of different tools” constantly being proposed, leaders are expected to help organisations cut through hype, and “how can you possibly sift through that if you don’t really understand the technology behind it?”
In his view, strong technical competence remains essential even at senior levels: “you can’t leave the technical skills behind, even if at a senior level it becomes a very different role”. But technical mastery on its own is insufficient.
The second, and equally critical, trait is influence. In an environment where “everything is changing so quickly” and options proliferate, progress depends on shaping decisions and building alignment. “If you can’t influence the decisions, you can’t get anything done,” Peter said. Peter frames influence not as authority, but as a mix of credibility and interpersonal skill. It requires “a bit of diplomacy”, ensuring you keep people onside, while still having the determination to push through outcomes that matter. Being influential means balancing empathy with resolve; knowing when to listen, but also when to insist.
For him, the leaders who make the biggest impact are those who combine genuine technical understanding with the personal effectiveness to steer choices and “get things over the line” when it counts.
What is the one non-traditional piece of advice that you would give to an AI or data leaders aspiring to get to do where you are at the moment?
Peter’s advice is rooted in personal resilience. Drawing on his scientific background, he reflects on spending years working on neural networks when it was “an unpopular field” and standing up and speaking at conferences to a room full of people where “basically, no one agrees with you”. That experience, he says, creates a form of self-confidence that is easy to overlook but increasingly essential for data and AI leaders: “the ability to not be afraid of being the unpopular person holding the unpopular opinion”. In today’s AI climate, that often means resisting fashion.
Peter is explicit that “not everything is an AI problem”, a view that “a couple of years ago was very unpopular”, when the dominant narrative was that “AI is going to happen and solve everything for us”. Holding that line requires the willingness to be “the sole voice in a room” even when consensus points elsewhere. For him, this is not contrarianism for its own sake, but judgement shaped by experience and knowing when enthusiasm has outpaced evidence. He traces this back to years of pursuing research that others “didn’t necessarily believe in”, which forced him to build a level of resilience that has become a major asset.
The lesson for aspiring leaders is clear. As technologies and trends shift at speed, leadership increasingly demands the courage to challenge prevailing assumptions, to say no when something is fashionable but wrong, and to stay the course when conviction is grounded in substance rather than popularity. Being effective, Peter argues, often means being comfortable with disagreement and accepting that influence does not always come from agreement, but from the confidence to stand apart when it matters.