MSc Data Science @ ETH Zurich | Computer Vision, LLMs, Anomaly Detection
Currently working on my thesis on Cognitive Diagnostic Models for LLM Evaluation and Skill-Aware Routing.
- Teaching Assistant — ETH AI Center: supervised anomaly detection projects (WinCLIP, zero-/few-shot visual inspection)
- Research Assistant — AIT Lab, ETH: pose estimation, 3D reconstruction, motion analysis
- Computer Vision Engineer — STROMA: real-time action recognition, pose detection (+15% accuracy)
| Project | Description | Stack |
|---|---|---|
| CroissantMiner | LLM-based Croissant metadata extraction from ML dataset papers — scaling to 100+ datasets | Python, TF-IDF, Mistral-7B |
| ETH Data Science Lab | WinCLIP-based anomaly detection for industrial visual inspection | Python, PyTorch |
| Plant Point Cloud Segmentation | Organ-level 3D segmentation for phenotyping — WACV 2023 Workshop | PyTorch, Open3D |
Python C++ PyTorch TensorFlow OpenCV scikit-learn Hugging Face Docker React Native