jan kompatscher

Ellis PhD-Student in Professor Antti Oulasvirta's Computational Behavior Lab at Aalto University, Finland. Working on simulation of human decision-making through online Reinforcement Learning. Interests lie in Machine Learning, Software Engineering, Explainable AI, and Visualization.

Education

• Ph.D. with Ellis program
  Aalto University Dec. 2024 - Dec. 2028
• M.Sc. in Human-Computer Interaction and Design
  Aalto University and University of Trento, Sept. 2022 - Oct. 2024
  120 ECTS GPA 97.66% (graduated with honours)
• B.Sc. in Visual Computing
  TU Wien, Sept. 2016 – Mar. 2021
  180 ECTS GPA 80.72%
• Exchange Program in Computer Science
  University of Calabria, Mar. 2018 – July 2018
  21 ECTS GPA 81.43%

Employment

• PhD-Student
  Aalto University, Computational Behavior Lab, Dec. 2024 - Present, Espoo, Finland
  Research on human modeling through RL
• Research Assistant
  Aalto University, Computational Behavior Lab, Nov. 2024 - Present, Espoo, Finland
  Research on visualization for RLHF
  Aalto University, Computational Behavior Lab, Mar. - Jul. 2023, Espoo, Finland
  Research on modeling human behavior when looking at information visualizations
• Developer
  EOS Solutions, Mar. 2021 - Feb. 2023, Oct. 2023 - Mar. 2024, Bozen, Italy
  Developing customizations of Microsoft Dynamics Business Central (ERP Software)
• Teacher (7th & 8th grade)
  Mittelschule Leo Santifaller Kastelruth, Dec. 2020 - Feb. 2021, Kastelruth, Italy
  Teacher of German, History and Geography
• Junior Developer
  Naramis, professional restaurant software, Jul. – Sep. 2018, Jul. – Sep. 2019, Völs am Schlern, Italy
  Migration of the software from one framework to another

Publications

• Kompatscher Jan, Danqing Shi, Giovanna Varni, Tino Weinkauf, Antti Oulasvirta. Interactive Group-
  wise Comparison Can Improve the Efficiency of RLHF. Proceedings of CGF 2025 (in preparation).
• Kompatscher Jan, Giovanna Varni. Interactive Groupwise
  Comparison for Faster Reinforcement Learning from Human
  Feedback. Master’s thesis, Aalto University, September 2024.
• Kompatscher Jan, Roy Alia Asiku, Filippo Gerin, Alessio Zeni. Detection of Glaciers and Equilibrium
  Line Altitude (ELA) for Trentino Province Using Remote Sensing Methods. Team 8, Public AI Chal-
  lenge 2023, December 2023. Mentored by Christian Casarotto, Francesca Bovolo, Milad Niroumand
  Jadidi.
• Kompatscher Jan, Hsiang-Yun Wu. Automatic Gradient-Preserving Stencilization of Raster Images. Bachelor’s thesis,
  TU Wien, March 2021.

Awards

• 1st Place, Public AI Challenge Trento - Developed an ML model for detecting glaciers and
  equilibrium line altitude (ELA), Public AI Challenge 2023.