About

As a PhD candidate at Stanford University, I work with Professor Stephen Boyd on modeling and solving convex (and nonconvex) optimization problems.

Before attending Stanford, I was a consultant and data scientist at McKinsey, where I worked on quantitative projects in the automotive and energy sectors. I received my MSc in Robotics, Systems and Control from ETH Zurich, where I was awarded the Willi Studer Prize for the best degree and the ETH Medal for my Master's thesis. I received my BSc in Mechanical and Process Engineering from TU Darmstadt, with the VDI Prize for the best degree.

Publications

M. Schaller and S. Boyd, "A Note on Optimal Product Pricing", arXiv preprint arXiv:2511.06156, 2025. Read paper

M. Schaller, D. Arnström, A. Bemporad, and S. Boyd, "Automatic Generation of Explicit Quadratic Programming Solvers", arXiv preprint arXiv:2506.11513, 2025. Read paper

M. Schaller, A. Bemporad, and S. Boyd, "Learning Parametric Convex Functions", arXiv preprint arXiv:2506.04183, 2025. Read paper

M. Schaller and S. Boyd, "Code Generation for Solving and Differentiating through Convex Optimization Problems", accepted for publication in Optimization and Engineering, 2025. Read paper

M. Schaller, G. Banjac, S. Diamond, A. Agrawal, B. Stellato, and S. Boyd, "Embedded Code Generation with CVXPY", IEEE Control Systems Letters 6, 2653-2658, 2022. Read paper

S. Löckel, S. Ju, M. Schaller, P. van Vliet, and J. Peters, "An Adaptive Human Driver Model for Realistic Race Car Simulations", IEEE Transactions on Systems, Man, and Cybernetics: Systems, 6718-6730, 2023. Read paper

P. Duhr, M. Schaller, L. Arzilli, A. Cerofolini, and C. Onder, "Time-Optimal Energy Management of the Formula 1 Power Unit with Active Battery Path Constraints", 2021 European Control Conference (ECC), 913-920, 2021. Read paper

P. Duhr, M. Schaller, L. Arzilli, A. Cerofolini, and C. Onder, "Analysis of Optimal Energy Management Strategies for the Hybrid Electric Formula 1 Car", 38th FISITA 2021 World Congress, 2021. Read paper

Software

CVXPYgen is a tool that generates custom solver implementations in C for convex optimization problem families modeled with CVXPY. It is suitable for embedded systems and also provides a Python wrapper for desktop applications.

LPCF is a framework for fitting a parametrized convex function to some given data, compatible with disciplined convex programming. This allows to fit a function that can be used in a convex optimization formulation, directly to observed or simulated data.

Teaching

I have been a teaching assistant for the following courses at Stanford and ETH Zurich:

EE 364A: Convex Optimization I. Professor Stephen Boyd. Stanford. Winter 2024-25, 2025-26. Created lecture slides on the convex-concave procedure.

ENGR 108: Introduction to Matrix Methods. Professor Stephen Boyd. Stanford. Autumn 2025.

151-0573-00: System Modeling. Professor Lino Guzzella. ETH Zurich. Autumn 2020-21.