Benjamin Kressig
Welcome! Glad you're here 🌿

I'm Benjamin, a MSc Student in Data Science at ETH Zürich focusing mainly on machine learning, LLMs and computational neuroscience. Gaining hands-on experience building LLM frameworks at Julius Bär and developing deep learning models at SLF has been a recent highlight. I also love building smart, user-focused applications, such as klavispiano.com, an interactive web app I created for piano practice.

Curriculum Vitae

Education
2023 — 2027
MSc Data Science
Feb 2024 — Jul 2024
2020 — 2023
BSc Electrical Engineering and Information Technology
Experience
Nov 2025 — Apr 2026
Machine Learning Engineering Intern

As part of the MLOps team, implement and optimize a framework for efficient deployment and inference of LLMs on a GPU cluster.

Sep 2025 — Oct 2025
Data Science Intern

During a 6 week internship project, developed a deep learning pipeline using PyTorch and U-Net models to automatically segment fractures in 3D point cloud data from mountain outcrops.

Feb 2023 — Dec 2025
Teaching Assistant
  • Course Computational Thinking: Preparing and holding weekly exercise sessions for 30+ undergraduate students on topics including algorithms, computational complexity and machine learning fundamentals. (Fall semesters 2024 & 2025)
  • Course Networks and Circuits Laboratory: Assisted in weekly sessions introducing 60+ students to experimental and measurement techniques in electronics. (Spring semester 2023)
Jun 2023 — Aug 2023
UAV Technician

Summer job assembling and testing new drones.

2015 — 2019
Electronics Technician Apprenticeship

Federal Diploma of Vocational Education and Training - Specialization in Embedded Software Development.

Projects

Klavis Piano

I developed Klavis Piano, a piano-practice web app that reads what you play on a digital piano over MIDI, follows you through the score note by note, and offers all kinds of helpful practice tools and visualizations. The landing page klavispiano.com covers exactly what it does, so it’s best to just explore it there :)

Computational Biomarker Discovery in EEG Data

In a team of three, developed a pipeline that uses Dynamic Causal Modeling to extract predictive features from EEG signals, enabling a machine learning classifier to distinguish a drug’s effect on brain connectivity.

Publications

Building AI Benchmarks from Exam Documents
Benjamin Kressig, Noah Koller, Imanol Schlag
2026 IEEE Swiss Conference on Data Science and AI (SDS)