We're PROJECT/THREE — a two-person founding team that has been on both sides of the table. One of us ran the wartime budget. The other engineered the platform.
We started the company because the defense industry keeps shipping the wrong answer: more closed, single-vendor platforms, priced for a world that no longer exists. The frontline doesn't need another silo — it needs the software layer that ties everything already on the ground into one coherent picture.
That's what we're building. Software first. Dual-use by default. Europe, but not only Europe.
Our software — an AI-driven data layer that ingests signals from any number of distributed sensors and resolves them into a single live position-fix. Hardware-agnostic. Publishes to any existing C-UAS stack.
Low-cost, mass-deployable RF sensor nodes. Parachute-ready, self-healing, edge-classified. Designed so the hardware is the cheapest part of the system.
Passive acoustic nodes — microphone arrays with on-device classification. Complementary to RF; works where RF doesn't. Deploys into the same fabric.
A proprietary, operator-validated corpus of drone RF and acoustic signatures — collected from live deployments. Used to train our detection models and licensed to partners training theirs.
The best defense systems of the next decade will be defined by their software. We build hardware only when it unlocks software that couldn't otherwise exist.
We publish standard interfaces before we build UI. Our software works with whatever's already on the ground — including systems we compete with.
Every feature is validated with operators in theater before it ships. If it doesn't survive a commander's stress test, it doesn't go in the build.
The same software runs emergency response, firefighting and infrastructure inspection. We don't rebuild — we extend.
No contact forms. No "request a demo" auto-responders. Email us directly — we read everything, and we answer.