We Patented How AI Reads a Building.
Our U.S. patent on AI-assisted engineering design is now issued. It covers the hybrid approach we've been building from the start, pairing AI that reads real architectural plans with deterministic engineering logic that enforces published code. Here's the work behind it, and what it unlocks next.
Today we can share something we've been working toward since the first line of code: the United States Patent and Trademark Office has issued our patent covering AI-assisted engineering design across building systems. Our platform turns a CAD floor plan into a validated, code-compliant fire sprinkler design and hydraulic analysis in minutes, and the method behind that is now protected.
Why this one was hard to get.
Plenty of software automates drafting. Very little of it can look at a real set of architectural plans and understand what it's looking at well enough to design a life-safety system from it. Real drawings come with inconsistent line weights, missing layer conventions, and a decade of drafting habits baked into every sheet.
That's the part that took years. Not the AI alone, and not the engineering rules alone, but the seam between them. Getting a patent on that seam meant proving it was genuinely new: that we weren't describing an obvious combination of "use AI" and "check the code book," but a specific architecture where each half does the thing it's actually good at.
The AI does interpretation. It reads unstructured drawings and turns them into structured engineering intent. The deterministic layer does compliance. Published standards and code requirements are applied as hard rules, not probabilities, not suggestions, not something a model inferred on a good day. When the drawing is ambiguous, the system doesn't guess. It surfaces the ambiguity for a human.
Articulating that boundary precisely enough to survive patent examination forced us to be more rigorous about it in the product, too. You can't claim a clean separation of concerns you haven't actually built.
What we were solving for.
The industry problem hasn't changed since we started. Demand for fire protection design keeps climbing. Codes keep getting more complex. The number of experienced designers and engineers is not keeping pace with either. Firms are asked to deliver faster, for less, without giving up an inch on safety. Most of the design process is still manual, repetitive, and dependent on expertise that takes a decade to develop.
Our position on that has never wavered, and our CEO Jason Tielve put it plainly in the announcement:
"Artificial intelligence shouldn't replace engineering judgment. It should amplify it. Our mission has always been to eliminate repetitive engineering work so professionals can spend more time solving complex problems."
That's not a hedge. It's an architectural commitment, and it's why the deterministic layer exists. In life safety, an engineer has to be able to see why a design is what it is. A system that can't show its work isn't usable here, no matter how good its output looks.
Why the claims aren't fire-specific.
We started with fire protection because it's where the pain is sharpest and where our team's expertise runs deepest. But the patent isn't written as a sprinkler patent. It covers AI-assisted engineering workflows across building systems, which means the same foundation extends to plumbing, mechanical, electrical, and the other disciplines that share the same underlying problem: read the building, apply the standard, produce a design an engineer can stand behind.
That was deliberate. We're not building a point solution. We're building an AI-native engineering platform for the built environment, and this patent is the technical floor it sits on.
Omar Hafez, who leads our AI innovation work, framed where this goes:
"This patent isn't the finish line. It's the foundation. We believe software should adapt to engineers. Our long-term vision is an AI-native engineering platform that understands buildings the way engineers do."
What's next.
The patent lands alongside real commercial momentum. Customer demand is growing, the platform is expanding, we've closed strategic funding, and we're continuing to invest in AI research. We're also in early licensing conversations in international markets, including exploratory discussions around deployment across Africa, where the shortage of engineering capacity is even more acute than it is here.
To every engineer, contractor, and designer who took a call with us, marked up our output, and told us exactly where we were wrong: this milestone has your fingerprints on it. Trust in this industry is earned drawing by drawing, and we intend to keep earning it.