NEOCAD files a patent application for AI technology that searches, verifies and modifies 3D CAD models. The application is pending grant.
Legnano (MI), June 22, 2026 – Italy is Europe’s second-largest manufacturing economy and the eighth worldwide: approximately 15% of national GDP, over 95% of exports and 50% of research and development spending*. In this context, NEOCAD brings AI to a key area of manufacturing: the technical office. Founded in late 2024, the startup has filed a patent application titled “System and Method for Processing, Searching, and Generating Parametric CAD Models Using Artificial Intelligence”. The technology is designed to autonomously search, verify and modify parametric CAD models within industrial design environments. “When discussing artificial intelligence, attention often focuses on major international players. Yet Italy possesses one of the world’s most important manufacturing heritages, built over decades of engineering, design, and production expertise. We believe that part of tomorrow’s AI innovation must originate right here, starting from the concrete needs of manufacturing companies. This patent represents our contribution to a technology developed in Italy to enhance Italian industrial know-how and make it even more competitive globally”, comments Luca Licciulli, CEO and Co-Founder of NEOCAD.
AI that operates on CAD models
The system is designed to generate and execute modelling operations in CAD software. Rather than only making suggestions, it works on the model, checks manufacturability and makes changes for production.
Making use of CAD archives
“Italian manufacturing companies hold millions of CAD models developed over the years. A wealth of technical knowledge that represents one of the company’s most important assets, but which often proves difficult to leverage”, emphasizes Licciulli. Fragmented archives and scattered information make it difficult to retrieve components and reuse designs. The technology described in the patent application searches by functional similarity: designers describe a component in natural language or provide a reference model to find relevant results in the company archive. “The goal is to transform the CAD heritage accumulated over the years into an intelligent and immediately accessible resource, reducing search times, design duplications, and loss of company know-how.”
From design to factory
The application describes a system that checks manufacturability against materials in stock, tools, technological constraints and regulatory requirements. When issues arise, the AI modifies the design and generates instructions to adapt it to the production context.
The goal is to reduce errors, rework and industrialisation time.
Closed-loop AI
According to NEOCAD, the distinctive feature of the technology described in the application is its ability to operate in a closed loop.
“Unlike most AI solutions, which merely suggest modifications or generate content to be validated later, the system is designed to act directly on the CAD model: it searches for components, verifies their manufacturability, and automatically generates the instructions necessary for modification”, continues Licciulli. The proprietary AI architecture is trained on mechanics and industrial design to understand physical and structural relationships between components. NEOCAD claims a reduction in geometric errors, from a 15–20% model invalidity rate to near zero. *Source for manufacturing figures: Industry Report 2025 (Confindustria Research Center).
