Leads the company's vision and the engineering of Show & Go — from the first experiments with failing public libraries to a platform running in daily production.
We started with a question nobody wanted to answer.
The question was simple: why does computer vision that dazzles in demonstrations quietly fail the moment it meets a real building?
We tested the available answers ourselves — public libraries, established frameworks, standard deep-learning approaches. The pattern repeated everywhere: strong results in controlled scenarios, breakdown in operational ones. Low-quality CCTV streams, poor lighting, crowded scenes, motion blur, real deployment scale — the conditions that define the actual world were precisely the conditions everyone optimised away.
So we built for those conditions instead. That decision became Show & Go, our vision infrastructure platform — and it set the character of the company: drawn to the problems others route around, committed to depth over breadth, and convinced that the hardest infrastructure is the most valuable kind.
"Most modern AI products can be replicated in days. We chose to build the things that can't."
The name carries the intent. Sirius is the brightest star in the night sky — a fixed point people have navigated by for millennia. That is the kind of company we are building: not the loudest in the room, but the most reliable point of reference in our field — the infrastructure others set their course by.
And like the kings and founders who began with nothing but conviction, we started without shortcuts — no borrowed models dressed up as products, no wrapper economics. Everything that matters here was built, tested, and earned in the field.
