Work Experience
2025 - Present
Machine Learning Engineer
Paymira AG
2024 - Present
Researcher
Hesiod, Zürich, Switzerland
Research-first company, trying to understand how neural networks "process" information. Current focus areas:
- VLM Interpretability: Studying reasoning capabilities and failure modes in vision-language models using sparse autoencoders for mechanistic interpretation.
- Representation Learning: Investigating geometric properties of representation spaces and emergent behaviors in large language models.
- Audio Diffusion Models: Developing latent diffusion models for video game audio generation with text and video frame alignment.
August 2022 - Sept 2024
Systems Specialist, Machine Learning
ETH Zürich, Institute of Robotics and Intelligent Systems
Contributed to a neuroscience project focused on processing of brain-computer interface data. Implemented data pipelines and machine learning models for neural signal analysis using Python, PyTorch, and cloud hyperscalers.
September 2019 - August 2020
Junior Data Scientist
ascarix AG (Avobis Group), Zollikon, Switzerland
Developed and deployed machine learning models for real estate data processing, including a probabilistic model for vendor data deduplication and an enhanced price estimation model using LightGBM and XGBoost. Implemented RESTful APIs with FastAPI and Docker, deployed on Microsoft Azure.
side projects
Jun 2016 - Jun 2019
Project Möbius
Built and maintained Möbius Code Editor, a data plotting environment for data analysis.
2010 - 2016
Android Development
things are heating up
Developed multiple Android applications with MVC architecture and RESTful API integration. Created some simple UI/UX visuals, wrote some simple Java code and for a split of a second (but not much longer) loved the Android SDK.
2011 - 2014
Game Development
Unity3D Projects
Developed Jump 'n' Run and 2.5D games for Android using Unity3D. Implemented pathfinding algorithms (A*, Dijkstra) and custom shaders.