Eric Vinchon is a product strategy leader with over 20 years of experience shaping the evolution of engineering software and 3D technologies. With a background in mechanical and software engineering, he connects product vision, business strategy, and technical execution across CAD and AEC ecosystems. Eric focuses on enabling digital threads, advancing enterprise 3D visualization strategies, and integrating AI into engineering workflows. He helps organizations rethink how complex engineering data is accessed, visualized, and leveraged to deliver scalable, intelligent, and connected solutions.
AI is transforming engineering, yet CAD data remains one of the hardest assets to leverage at scale. This session explores how geometry-aware architectures and embedding techniques allow machines to understand shape directly, enabling similarity search, prediction, and business intelligence across engineering workflows.
AI is rapidly reshaping engineering workflows, yet CAD data remains one of the most complex and underutilized assets in machine learning. Unlike text or images, CAD models are multi-modal, hierarchical, and deeply structured, making traditional AI approaches difficult to apply at scale.
This session explores a geometry-aware approach to machine learning for 3D data. We’ll examine how supervised architectures can classify parts and features, why scaling them requires more than labeled datasets, and how embedding-based techniques enable machines to learn shape directly.
By transforming geometry into machine-readable representations, engineering teams can unlock similarity search, clustering, and predictive insights, correlating 3D data with manufacturability, cost, sourcing, and quality outcomes.
The focus is not on hype, but on what is happening now, and what organizations must prepare for next.