Machine Learning Intern (Engineer / Scientist)

Builtstop is a startup developing smart earthquake protection solutions for buildings. By combining advanced technology with practical design, Builtstop aims to improve resilience and protection in urban environments. Technical expertise and creative thinking come together to address real-world challenges.Role description: You will work directly with the technical team on using Machine Learning methods (such as GNNs) to learn from structural models and predict suitable retrofit configurations for different building typologies. You will work and contribute directly to a project that connects structural engineering, simulation and machine learning.Specifically, you will:Prototype and test ML architectures for representing and learning from building structures.Develop models to predict optimal retrofit strategies.Generate and process datasets from structural simulations.Benchmark ML predictions against physics-based results.Explore surrogate modelling and optimization approaches to reduce computational cost.QualificationsYou have hands-on experience with machine learning and can move from an idea to a working prototype independently.You have experience or strong interest in Graph Neural Networks / geometric deep learning.You are fluent in Python and familiar with modern ML frameworks such as PyTorch, PyTorch Geometric, JAX, or similar.You are comfortable working with scientific or engineering datasets and numerical simulations.You enjoy research-style problems where the solution is not predefined.You are self-directed, curious, and comfortable testing ideas quickly.You are fluent in English.Bonus: experience with structural mechanics, finite-element modelling, surrogate modelling, optimization, or physics-informed machine learning.Essential: You're already based in (or near) Zurich.How to applySend a short note on why this role fits you, plus your CV, to info@builtstop.com. We review applications on a rolling basis.