Søren Taverniers
Research Scientist, Mechanical Engineering
Søren works at the intersection of AI and scientific workflows, leveraging machine learning and agentic frameworks to accelerate and adaptively automate existing pipelines for uncertainty quantification and optimization.
He is a former member of the Tchelepi research group at Stanford Energy Science & Engineering, where he worked on accelerating risk assessment of induced seismicity in subsurface problems through multi-fidelity Monte Carlo-based frameworks combining physics-based simulation in GEOS with data-driven methods including graph neural networks.
He has subject-matter expertise in uncertainty quantification, machine learning (including operator learning and graph neural networks), and multi-physics modeling, and has published 10+ journal papers in high-impact venues spanning computational physics and numerical methods.