We build AI systems that automate, optimise, and scale mechanical design. From geometry engines to full design pipelines: built on research, deployed in production.
Component AI is founded by top experts from the Danish mechanical design industry and the AI research community, with the aim of developing algorithms and machine learning technology for processing of geometric data in the design industry.
We use this technology to solve problems in mechanical design in tight collaboration with clients, applying geometric algorithms and artificial intelligence to optimise design processes and reduce manual work — from initial design sketch to finished, manufacturable part.
Applying methodology close to the research state-of-the-art, Component AI has tight links with the academic AI community in Copenhagen. We develop algorithms that push the current research frontier, and we hire top international graduates in machine learning, AI, computer science, and mathematics.
If you have a hard technical background and are interested in building AI that solves real-world problems, please get in touch.
We work with manufacturing companies that want to move faster, scale design quality, and iterate faster with reduced manual iteration. Our work ranges from single-project engagements to full pipeline integrations.
End-to-end AI pipelines that take a component from initial geometry to manufacture-ready output — automatically. Optimisation, DFM, documentation, and export in one continuous flow. Built for teams running high volumes or complex product families.
Point-by-point optimisation of 3D geometry for manufacturing constraints, structural performance, and multi-physics requirements. We compute the geometry that actually performs, not the one that was easiest to draw. Injection moulding, CNC, and 3D print.
Automated design-for-manufacturing analysis and correction. We replace weeks of manual DFM iteration with algorithms that identify, quantify, and fix manufacturability issues across shrinkage, warpage, draft angles, wall thickness, and flow paths — simultaneously.
AI-generated technical documentation from geometry and simulation outputs. Tolerance specifications, DFM reports, drawing annotations, and quality documentation produced automatically — consistent, traceable, and always aligned with the current geometry.
Computer vision systems for in-production quality inspection. Custom models trained on your part geometries and defect profiles — detecting deviations, surface defects, and assembly issues at production speed, without manual inspection bottlenecks.
CoDesigner is our commercial product for development and manufacturing teams. It takes a design intent and existing model and produces optimal geometries — geometry computed by physics, optimised at every surface point, manufacture-ready. Built on the Component AI geometry engine and years of production-validated engineering knowledge. We are opening access to a first group of teams now.
Component AI is built by Component Group — a Copenhagen-based consortium of Component 2.0, a product and engineering consultancy, and Component AI, a team of leading specialists in AI for engineering. Over more than two decades, the group has trained and educated more than 4,000 engineers. We have seen first-hand how exceptional they are at creating concepts and starting geometries — and how they spend the majority of their time on DFM corrections, tolerance dimensioning, and manufacturing iteration. That is what we automate.
The geometry engine is built on state-of-the-art shape analysis methods that parametrise and quantify smooth geometry transformations from Riemannian metrics on spaces of non-linear transformations. Coupled with language-to-geometry generators and end-to-end differentiable numerical schemes, it optimises all points simultaneously for any criteria — in scalable geometry generation pipelines. Years of engineers, experts, and researchers developing methods in algorithmic mechanics and algorithmic mathematics enter the optimisation process, allowing a body of knowledge previously locked in human expertise to become operational algorithmically.
With AI, we can flip the paradigm: stop analysing what already exists, and start designing what should exist. Geometry in code supercharges computation. CNC already machines point by point. AI unlocks the geometry. This is Computational Numerical Design — the design equivalent of what CNC did for manufacturing. A complete break from manually drawn lines, towards physics-driven geometry computed from the ground up.
Component AI was founded in 2024 by Janus Juul Rasmussen and Stefan Sommer.
Professor of Computer Science at the University of Copenhagen and Head of the Pioneer AI section at DIKU; founder of Sommer AI. Leads Component AI's technical vision and development strategy.
Founder of Component 2.0, a leading Danish mechanical design consultancy. Leads overall strategy and business development, bringing deep industry expertise in mechanical design and manufacturing.
Stay informed about our latest breakthroughs, partnerships, and innovations in AI-driven design automation.
We like to push boundaries at Component AI. Behind the CFD simulation is the core of what we do: tightly coupled geometry and physics computation.
dxdydz transformations compensating geometry for manufacturing deviations. We predict transformations for millions of nodes in the geometry simultaneously.
Blending design for manufacturing with algorithmic thinking, mathematics, statistics, software programming, and AI is now producing tangible component breakthroughs.
Whether you have a specific project in mind or want to explore what is possible, we would love to hear from you.
If this vision resonates with you, come join us. We are a small, ambitious team working at the intersection of physics, geometry, and machine learning — solving problems that haven't been solved before. We have offices in central Copenhagen and Hørsholm.
career@component-ai.dk →