Luis Kaiser

I am a PhD student in Computational Mathematics and AI at the Oden Institute, The University of Texas at Austin. I currently design autonomous systems, including predictive planning, robotics and foundation models grounded in math advised by Richard Tsai.

I completed my B.Sc. and M.Sc. in Mathematics at Technical University of Munich (TUM) and University of Wuerzburg. For my master's thesis, I developed multi-step gradient descent for deep physical modeling advised by Christian Klingenberg. For my bachelor's thesis, I created a method to stabilize policy convergence of a deep reinforcement agent from sparse, stochastic rewards together with Andreas Hotho.

I was also a Software Engineer at Bloomberg in London where I built large-scale AI research and forecasting systems.

lkaiser (at) utexas.edu
Google Scholar, LinkedIn

Publications and Updates

Efficient Numerical Wave Propagation Enhanced By An End-to-End Deep Learning Model teaser

Efficient Numerical Wave Propagation Enhanced By An End-to-End Deep Learning Model Luis Kaiser, Richard Tsai, Christian Klingenberg. ENUMATH23, Springer Nature, 2024.

paper, code, arXiv

Starting my PhD in AI at The University of Texas at Austin

Joining the Oden Institute for Computational Engineering and Sciences as a PhD student in Fall 2024.

Joining Bloomberg as a Software Engineer

Working on a large-scale AI forecasting system in London, UK starting in Feb 2023

Completed Google's Foobar Challenge

Google's invite-only algorithmic challenge involves solving a series of discrete math puzzles using Python.

Autonomous Multi-Agent Deep Reinforcement Learning for Ice Hockey teaser

Autonomous Multi-Agent Deep Reinforcement Learning for Ice Hockey Luis Kaiser, Pedro Silva, Austin Aurelio, Marlowe Johnson. RL Computer Vision Competition, 2021.

paper, code

Evaluation and Improvement of Deep Reinforcement Learning Agents teaser

Evaluation and Improvement of Deep Reinforcement Learning Agents Luis Kaiser, Julian Tritscher, Padraig Davidson, Andreas Hotho. Bachelor Thesis at University of Wuerzburg, 2021.

paper, code

Show All Publications