Ahmad Alkadri’s route into software engineering ran through wood acoustics, building physics, and a factory floor in Luxembourg. Now, as Technical Product Lead at Tektome, he is working on a harder problem: teaching an AI agent to understand the space a building occupies.
A path into software, walked rather than planned.
Ahmad’s background spans research, civil engineering, simulation, and software development. He is quick to point out that software engineering was never a formal destination; it was somewhere he arrived by doing the work.
The habit started in academia. Through his Master’s and his PhD in France, he was already writing software: simple statistical scripts at first, then a tool to observe and calculate fissure area. He learned in practice, little by little, and the formal classes he took filled in the theory behind what he had already picked up by hand. That mix, ground-level practice supported by academic pillars, is what eventually carried him into the profession.
When a single bug stops the factory.
After his PhD, Ahmad joined a startup in Luxembourg as a simulation engineer, doing the thermal and mechanical calculation work typical of civil engineering. One thing led to another, and he ended up on the company’s principal ERP tool: the system that automated manufacture and assembly, wired directly into the robotics that produced pre-fabricated houses.
That was a different kind of engineering. Writing tools for yourself, or for a handful of colleagues on their own machines, is one thing. Deploying something factory-scale, for clients, running live production, is another.
Key takeaway
Deploying at factory scale changes the stakes. A tool you write for yourself can fail quietly; one running an assembly line cannot. A single bug can halt production, and every halted hour has a cost.
He felt that cost directly. The assembly lines started earlier than the engineers, so more than once he was woken in the small hours, told a bug had stopped production, and had to get to the factory to fix it. By then an hour or more of production was already lost. Reliability, in that setting, was not an abstraction.
Why Tektome, and why now.
Ask why he chose Tektome and Ahmad answers with a grin (partly, he admits, because it is in Japan). The honest reason runs deeper. He had been looking for a way into agentic software engineering for a while, convinced that the field is moving towards an era in which people and AI systems work side by side to solve problems.
Underneath that is a throughline he traces across his whole career, from research to construction to house assembly: he wants his work to actually help people. Tektome, he felt, could take him closer to that.
Teaching an agent to see in 3D.
The technical problem he finds most interesting at Tektome is also one of the hardest in the field: making an AI agent spatially aware.
Agents are strong at reading. They handle text well. What they struggle with is space, and space is exactly what AEC runs on: objects behind other objects, objects wedged between others, pipes snaking in and out of walls. Much of that work is repetitive, which is precisely the kind of thing that should be a candidate for automation, or for handing to an agent. The catch is that agents are still poor at reasoning about three dimensions.
Agents are great at reading and understanding text. They’re still not good at spatial analysis. We need to give them the tools to “see” around, to work in three dimensions.
Ahmad Alkadri, Technical Product Lead, Tektome
That is the work: building the spatial and computational tools an agent needs to perceive its surroundings, so it can take on the repeatable 3D tasks engineers currently do by hand.
Everyone is a product engineer now.
Ahmad’s title has moved from Software Engineer to Team Lead to Technical Product Lead, though at Tektome, he jokes, everyone is a product engineer now. The progression felt natural: working alone on tools, then working inside a team under a lead and seniors on larger ones, then leading engineers himself, then helping shape the product directly.
Each step widened the view. Product work meant understanding the platform more deeply, engineering included, and getting closer to what clients actually need and which problems are worth solving. These days the pace is steadier: Tektome has sharpened its focus to a single platform-product with sub-applications inside it, and he works on some of those sub-apps.
Cheaper, faster, and built to be lived in.
What excites Ahmad about the future of software, AI, and BIM in AEC is speed, and, if it can be managed, lower cost. As AI automates more of the work and digital transformation continues, he expects the cost of building or obtaining a home to fall, and that, to him, is the point.
Housing should be lived in, not treated as a speculative tool. As needs grow, we should be able to supply more of it, in better and more efficient ways.
Ahmad Alkadri, Technical Product Lead, Tektome
It is a fitting ambition for someone who arrived at software by way of wood, buildings, and factories: technology as something that lowers the barrier to a basic human need, rather than an end in itself.
Join the team
If Ahmad’s work sounds like the kind of problem you’d want to spend your time on, spatial reasoning, agents, and 3D at real scale, take a look at where you might fit on the team.