When you first meet Ismail Negm, you don’t immediately guess that he has a background in quantitative trading, war photography, cave diving, and electronic music. But by the time you’ve talked for 10 minutes, you understand: he thrives at the edge of complexity.
Ismail is PhysicsX’s Director of Technology, a role that encompasses software engineering, cloud engineering, IT, and infosec. It's a lot — but if anyone’s built for it, it’s him.
We sat down with Ismail to talk about his path, the kinds of problems he’s tackling today, and how he's shaping and scaling his team.
Let’s start with your journey. How did you get into this space, and why?
I’ve always been around computers, I got my first one when I was four. My dad was an electrical engineer, so our house had a rotating cast of early machines — Apple II, C64, Tandy, and Atari. I assembled my first computer at eight and learned to code BASIC in elementary school in the ’80s.
I was planning on going to med school, motivated by a program around robotics and telesurgery. I studied behavioral neuroscience and was encouraged into pure mathematics by a computer science professor. This was during the AI winter, and by pursuing a combination of neuroscience and math, I thought I could get closer to symbolic AI and computational neuroscience.
My first real job out of university was as an options trader at HSBC. I had a small team, and we were doing multi-leg options trades, tracking their spreads manually, and trading on terminal-based user interfaces. I wrote small applications to help track our trades, which saved time but also raised eyebrows in IT. When I realized how slow the vetting process would be, I left to start my first company: Tiamat Trading.
Tiamat was born during a weekend hackathon backed by VCs, where a friend and I implemented Benoit Mandelbrot’s Multi-Fractal Model of Asset Returns in the J programming language. We won the people’s prize and came second with the judges, who said they hadn’t quite understood what we’d built... but had still decided to give us the main prize, which was a one-hour meeting with the VCs. I quit my job and went all in. We developed the model into a simulation framework for testing trading strategies and for options pricing, and we developed a column-oriented database, better suited to storing time series. We weren’t successful, but it was a very formative experience.
I then moved to Egypt and started my second company, a pivot on the first. We raised enough to build a team of 30 engineers, and it took off quickly. We developed market risk management software for the central bank, the country’s largest retirement fund, and the region’s largest brokerage, to cite a few. Revenues were very strong until everything collapsed during the Arab Spring in 2011. So I took a sabbatical... and went to Libya as a war photographer for London News Pictures.
What pulled you back into corporate work after that?
Up until that point, I’d been either trading or trying to get a startup off the ground. I was 31 and felt like I needed experience in a more structured environment where I could focus on my craft as a quant and software engineer.
I went back to Montreal and worked for hedge funds where I experimented with FPGAs, developed with early-days CUDA, and built risk and trading infrastructures. I worked for a technology startup as Director of R&D, building their recommender systems and a point-in-time database. I eventually ended up at McKinsey as a Staff Software Engineer and was quickly promoted to lead several product teams as an Associate Partner and Director of Engineering.
How did you end up at PhysicsX?
I met Jacomo at McKinsey, and when I looked into PhysicsX, I was immediately interested in what the company was building. I was introduced to others at PhysicsX, and they each left a very strong impression on me. I instantly liked everyone I spoke to. It was that combination: the people, the mission, and the fact that this felt like actual deep tech.
There’s a lot of talk in the industry about “deep tech,” but most of it is either still decades from market (e.g., quantum computing) or chasing hype cycles. PX is different, it’s not a gimmick. We’re solving engineering problems that are here now, and we’re doing it in a market that’s already hungry for the solutions. That’s a rare and exciting combination.
Now that I’m in the position of speaking with candidates, I can see them light up when I tell them what we’re working on. It’s not smoke and mirrors. It’s frontier work.
What does your day-to-day look like?
I’m responsible for software engineering, cloud engineering, IT, and infosec, so my day involves a lot of context switching. Right now, my focus is on building effective teams and strengthening our foundation for scale. Sometimes it’s orchestrating design sessions, sometimes it’s finding the right hires or vendors, or just making the right call quickly so we can keep momentum.
I have to think both tactically and strategically: Where are we going? How do our technical decisions position us for partnerships? How do we build credibility so people don’t look at our age as a startup and assume we can’t deliver at enterprise scale?
How is PhysicsX different from other players in the space?
Our platform enables high-fidelity multi-physics simulation through AI inference across the entire engineering lifecycle, and that is the differentiator.
Most engineering systems are built for a single stage: design, simulation, or manufacturing. We integrate across the entire lifecycle. But — and this is key — we don’t aim to replace the legacy tools. We interoperate with them. Our customers have used those tools for decades. We respect that, and we meet them where they are while significantly augmenting their capabilities.
What kind of people thrive here?
We expect a lot.
We aren’t a tech giant where there are numerous layers of automation built by thousands of engineers between an engineer and production infra. We don’t have an in-house version-control system managing billions of lines of code. We haven’t built a golden path for every deployment or upgrade. Our engineers need to get involved at all stages of the software development lifecycle.
We aren’t a low-stakes web application company with a high risk tolerance and a low cost of failure. Our customers are large enterprises building physical products with highly valuable IP. They have strict security and compliance requirements and expect our software to be enterprise-grade.
To thrive as a software engineer at PhysicsX, you have to own your craft across the software development lifecycle, from gathering requirements to monitoring production workloads. We value a strong sense of ownership and the skills to deliver. We put a lot of effort into helping our software engineers grow their skill sets, so those passionate about their craft will find PX an excellent environment for growth with plenty of opportunities to put their skills to the test.
What are the big technical challenges PX will address in the next 3–5 years to deliver maximum value to customers?
There are many, but let me give you a flavor.
In typical ML, you’ve got terabytes of tabular data — every sample is 2MB in size. In our world, it’s terabytes of geometric data, with each sample being around 800MB. So now, you’re dealing with massive, deeply structured data that you need to serialize, compress, and think about a common representation for across multiple proprietary formats.
Another challenge is the orchestration of simulation workloads: orchestrating hundreds of simulations, each taking hours to run, scaling the distributed infrastructure, while remaining within license constraints. Traditional engineering software is not designed to be cloud-ready, and yet we need to run it like it is.
Finally, writing custom compilers for our in-house model architectures. It’s how we get to real-time optimization, with real-time multi-physics inference while morphing a geometry. These are just some challenges that are top of mind.
What do you do to recharge outside of work?
Too many things! I’m a technical scuba diving instructor, I’ve done war photography, I do sound design, and create experimental electronic music with a modular synthesizer I’ve built (you can get some very interesting sounds out of it).
Music is probably the most stable of my obsessions (a lot of it is programming-based), though the others come in waves. I like to do concerts in domes with 100+ speakers, pitch black, fully immersive sound experiences, using software I’ve written to distribute the audio across the different channels. Once you’re a programmer, everything becomes a programming problem.
An example of inspiration for my music is George Friedrich Haas’ In Vain. It's a 40-minute orchestral composition playing on the tension between traditional harmony in equal temperament and spectrum-based tuning. It’s not the kind of music you could just play in the background; best to put your headphones on and sit there for 40 minutes. You’ll either love it or hate it.
What advice would you give to someone just starting out who sees the world like you do?
Stay that way. Seek out work that fuels your passion for engineering. Some projects you can hack on over a weekend, but others require the kind of scale, infrastructure, and compute power that only the right company can provide.
Bring your enthusiasm to work, and choose a place that amplifies it. If you ever find yourself stuck, uninspired, or not growing — leave. Your passion is the essential fuel. Technology evolves too quickly for complacency; what you’re good at today might be irrelevant tomorrow. To stay ahead, you need to be passionate and willing to put in the time to learn things on your own. Work needs to be a conduit for that passion, else it's not worth doing.
Blitz Round
Mac, Windows, or Linux? Mac now, having lived most of my life on Linux.
Favorite keyboard shortcut? Command + space. Very meta and keeps my hands on the keyboard.
Tabs or spaces? Spaces (let’s not start a war).
Slack message or huddle? Huddle.
Broken Wi-Fi or low battery? Low battery. I’d rather not have Wi-Fi when I’m coding.
One piece of tech you won’t upgrade? My Raspberry Pi Kubernetes cluster. I like the fact that it's got limited memory and CPU.
Dream hackathon team, dead or alive? Peter Norvig, Leslie Lamport, Arthur Whitney.
Favorite book or podcast? Signals and Threads for podcast, and Team Topologies by Manuel Pais and Matthew Skelton. But it’s actually impossible to choose just one book, so the answer may be different next time.
Go-to coffee order? Double espresso. One or two. Three if it’s one of those days.
What makes a candidate stand out? 1) Side projects. 2) Depth of passion. 3) Breadth of curiosity.
At PhysicsX, we solve hard problems. People like Ismail make it possible. And if you're someone who sees the world the way he does — someone who wants to build, solve, and push beyond traditional engineering paradigms — we're hiring.