Back to jobs
S

Junior Computational Geometry Engineer

SaekiZurich6 days ago
Full-timeVia Arbeitnow
Apply now

About this role

SAEKI - Manufacturing at software speed Europe's defense industrial base cannot respond at the speed modern conflict requires. Legacy suppliers operate on month-long timelines. Tribal knowledge is retiring. Supply chains are fragile. SAEKI is building the operating system for autonomous factories - the infrastructure that lets Europe manufacture mission-critical components at software speed. The Role As a Junior Computational Geometry Engineer, you work on the geometry core of our Manufacturing Intelligence stack: the engine that takes an arbitrary customer CAD model, reasons about how to build it, and then proves that we built it right. You start on well-scoped parts of that engine, next to the engineers who built it, and grow into owning them. Your code runs in the factory. What We're Looking For Foundations . Degree in computer science, mathematics, mechanical engineering or a related field, with real depth in geometry, linear algebra and numerics. Internships, thesis work or serious side projects count. We do not expect years of production experience. B-rep Basics . You are comfortable with topology, NURBS, meshes and point clouds. You may never have shipped a robust boolean, and you still understand why floating-point robustness destroys naive implementations, and you want to learn how the field handles it. Kernel Exposure . You have worked with a geometry kernel and mesh libraries in at least one project: OpenCASCADE, CGAL, libigl, Open3D, PCL, Parasolid or comparable. Technical Ability . Solid modern C++ (17+) plus Python for tooling and rapid iteration. Rust is welcome. You write code others can read, test and maintain. Manufacturing Curiosity . You want to understand what happens once geometry meets a spindle, a nozzle or a fixture. CNC, additive or metrology exposure is a bonus, never a requirement. First Principles Thinker . You reason by physics and mathematics, not analogy. When an algorithm fails on a real part, you dig until you know whether the bug s