At PhysicsX, we work at the edge of what’s possible in simulation and AI. For Catriona Stokes, a Simulation Engineer on our Delivery team, it’s that mix of deep physics, creativity, and impact that makes the work meaningful. From cardiovascular flows to aerodynamics to the science of a beer glass, Catriona brings an intuitive, cross-disciplinary mindset to every project. We sat down with her to talk about her journey, what it’s like working on the Delivery team, and how she sees AI changing the game for engineers.
You started out more into the arts than the sciences. What changed?
It’s true, I grew up in a creative family surrounded by people doing music, languages, and arts. Science was something I enjoyed, but wasn’t something I initially saw myself pursuing long-term. That changed when I was around 15 or 16, after we moved from the US to the UK. I had some fantastic physics teachers at school and found myself drawn to the mindset that the subject encouraged, so I chose to pursue a degree in physics and mathematics.
In my final year at uni, I took a course on fluid dynamics — it was intense, challenging, but really fascinating. I remember thinking, this is everywhere. There are just so many applications to it and the math is really interesting and kind of mysterious.
I knew that I wanted to get out of the purely theoretical world and into one where I could use science in a tangible way. That’s what led me to a one-year master’s in computational fluid dynamics. I loved it and I knew I wanted to keep going in that direction. I started my career at Dyson as an aerodynamics engineer, then had a brief stint as a contractor in aerospace. I didn't really love that and felt a bit lost, so I pivoted to a PhD in cardiovascular flows at UCL which combined my lifelong fascination with medicine, interest in learning more about system-level dynamics, and background in CFD. It was a great experience which reignited my excitement in the field.
The simulation space is always evolving, creating new possibilities and applications in areas like digital twins and surrogate modeling — the potential for that is huge. I think that I enjoy simulation so much because it's not just a technical role. You have to use intuition and creativity and work in interdisciplinary teams in order to unlock maximum value from your modeling.
What brought you to PhysicsX?
After finishing my PhD, I was doing a postdoc in the same research group. It gave me time to work on some new research questions while figuring out whether I wanted to stay in academia or move back into industry. I missed the pace and excitement, but I didn’t want to be limited to one sector, and many roles in simulation are deep into one application: F1, turbines, vacuum cleaners, and so on.
That’s when I came across PhysicsX, where you can be working across semiconductors, energy, aerospace & defense, materials, automotive — it’s not boxed in. The thing that interested me most was the aim of combining simulation with AI and developing rapid optimization techniques. That’s where engineering is headed, and I could immediately see how valuable the tools we are building would have been in all my past work.
What’s a typical day like on the Delivery team?
Mornings are golden — I get a few focused hours to set up simulations, do post-processing, and build slides before our US teammates come online. Then in the afternoon, we have project stand-ups to catch up on project progress and next steps, and cross-functional meetings to learn about the work happening across PhysicsX.
Our customers are in many different time zones, so we usually have our customer reviews late in the afternoon. We typically meet with them two to three times a week, discussing our modeling work, planning the next phase of what we're doing, and strategizing. We work collaboratively, often co-engineering solutions with our customers, so it's important that we're in constant communication and well-aligned.
Every few months, we also go on-site. You see the physical systems you’re simulating, talk to the engineers designing and testing the systems, and discuss your work with the wider team. At its core, our work is about modeling real-world systems, continuously improving the accuracy and fidelity of those models, and above all providing valuable insight to our customers that would not be possible without simulation. Being on-site really enhances the sense that your work contributes to building something real, as data can sometimes feel abstract in isolation.
You’ve also modeled… beer temperatures? Tell us more.
(Laughs) It started as a post-work pub debate. One of our data scientists had a theory that drinking beer in half-pints keeps it colder on average than drinking a full pint — less time for it to warm up in your hand. Naturally, we started discussing the science behind it, working out the surface area to volume ratio of the glass. And by the time we got to the fundamental equations of fluid dynamics, we realized that this was a bit beyond back-of-the-envelope math.
It turned into a mini project where I built a model in Python to simulate heat transfer in beer glasses under different conditions: hand warmth, outside temperature, beer temperature, and so on. The model ended up being quite complex, and it turns out the original theory does hold under most scenarios.
Funny thing is, I found myself using similar modeling logic just last week on a customer project!
P.S. We’ll be sharing the details of the "beer model" soon on Medium. Stay tuned!
What’s the biggest challenge in your role?
You’re almost always working on something new, usually involving systems with complex and highly non-linear behavior, and with strong coupling between system components. These problems typically require a variety of multiphysics modeling approaches. You’ve got to be comfortable learning on the fly, asking the right questions, and finding ways to leverage time and compute resources efficiently.
Also, we’re not working in isolation. Simulation engineers at PX work alongside machine learning (ML) engineers and data scientists, and we all work closely with our customers’ teams. So you need to speak each other’s language and translate insights into something actionable. Cross-discipline collaboration is essential and you need to be a team player.
What simulation trends are you excited about over the next five-year horizon?
Naturally, AI — that is the next big thing in engineering. Large language models (LLMs) are already allowing us to work at a much faster rate because we can use tools like ChatGPT or Claude to build macros or write post-processing scripts.
Beyond that, much broader transformations are coming from ML, unlocking rapid simulation and optimization and enabling us to create digital twins across nearly every area of engineering. It’s revolutionizing how we design, iterate, and validate systems. And it frees up our time and energy to think in more innovative, strategic ways.
We’re also seeing simulation software platforms increasingly leverage GPU architectures, which are enabling more complex simulations to run much faster. The ability to model complex multiphysics systems is always improving — and the pace is accelerating, compounding with the advancements in AI/ML.
Cloud-based computing is another major shift. It’s making simulation far more accessible because engineers don’t need on-premises HPC facilities to run high-fidelity models. That democratization of simulation is huge.
The future of simulation is intelligent, agentic, and integrated — I think that’s going to completely transform how we design and engineer systems.
AI — substitute or a co-pilot?
Exactly, it’s a co-pilot. You still need engineering judgment and experience to generate high-quality data to train the models. If your data isn’t validated and grounded in reality, even the most advanced ML models won’t give you anything useful.
So the value comes from combining domain expertise with AI, not replacing one with the other.
How do you reset and get inspired?
Most weekdays when I'm not at work, I go to the gym to lift weights. It’s very grounding after a long day of thinking, and also humbling. It teaches patience and consistency, which mirrors a lot of what we do in engineering — iterating and leveraging past efforts to build capability over time.
I also love photography, archaeology, and ancient history — I'm fascinated by the Neolithic, Bronze and Iron Age periods — the mystery around them is so intriguing. When I take time off, I often go on long walks looking for stone circles. Roman engineering is another one that amazes me. It’s incredible to think how many of the principles we rely on today were laid down thousands of years ago.
Advice to someone starting out in simulation?
I’d say for people early in that career, it’s really important to refine their technical foundation, but something to start thinking about, developing, and demonstrating early, is your soft skills.
If you’ve finished a degree in engineering or sciences, you’ve already shown that you can learn. You can always pick up new technical skills over time. But you also need to develop things like engineering judgment, creative problem-solving, presentation, and communication. Whether you’re working with customers or presenting internally, it’s all storytelling. You need to identify what matters and learn how to frame your conclusions in a way your audience can relate to, take forward, and apply.
Another thing that really accelerated my own growth was thinking in systems. It’s easy to get stuck focusing on one small component — maybe that’s your task. But greater insight is available when you understand and integrate how that component fits into the whole system. The more you push yourself to see that bigger picture, the more impactful your work will be.
Last question — book recommendation?
Thinking in Systems by Donella Meadows. It's about how everything connects — engineering, society, nature. You can apply that to engineering systems or just, you know, the world and the universe. I think it's a brilliant book.
We're growing our Delivery team! If you want to help build AI that is accelerating hardware innovation across advanced industries, explore our open roles here.