Translate hand gestures/signs into precise CAD/DCC commands in real time.
🔓 Project is open sourceMotionCoder is an AI-powered, modular, on-prem framework that maps 3D hand gestures/signs to intents + parameters, live values (e.g., draw circle, split edge, move 5 mm) and executes them in target apps. Beginner-friendly core gestures; no sign-language knowledge required.
Benefits
Real-time, low-latency pipeline • Undo-safe integrations • Reproducible runs
No videos yet — the project is still in an early prototyping phase. A first demo is in progress and will be published here once ready, with regular updates on YouTube. In the meantime, you can read the pitch below.
Short overview of MotionCoder: problem → solution and why
Note: I am not looking for investors. My primary interest is in developers and engineering partners.
Originally, I wanted a very short domain name (for example, xxxx.ai). Since most four-letter .ai domains are already taken, I decided on xtan.ai – it happened to still be available, so I registered it.
The logo is based on a straight line and a tangent curve: together, they form an
X with an integrated S-shaped arc. You can visualize this by plotting
y = 1.5x and
y = tan(-0.5x) { -π < x < π }.
In short: I’m moving into the software development field.
More about me: xcnc.eu/company
Email: info@xtan.ai
GitHub: xtanai/overview