Marco Voegeli

Machine Learning Enthusiast, Data Science Aficionado, and Occasional Coffee Addict

About Me

Hi, I'm Marco – a Computer Science Master graduate specialized in machine learning, data science, and information security. With experiences spanning Switzerland, Romania, Poland, Indonesia, and Turkey, I bring adaptability and a global perspective to any team. I don’t do fluff—I build solutions and drive the future.

Education

MSc. Computer Science — ETH Zürich

Major: Machine Intelligence, Minor: Information Security

Core Courses: Natural Language Processing, Probabilistic AI, Advanced Machine Learning, Deep Learning

Thesis: Engineered a machine learning algorithm to predict nocturnal hypoglycemia in children with type 1 diabetes (AUROC: 0.78) using a limited dataset.

ICLR 2025 Paper

Duration: 09/21 - 07/24

BSc. Computer Science — EPFL, Lausanne

Core Courses: Algorithms, Probabilities and Statistics, Parallelism and Concurrency, Object-Oriented Programming

Duration: 09/18 - 07/21

GCE A Levels — MEFIS, Izmir

Core Courses & Grades: Math (A), Physics (A), Chemistry (A)

Duration: 09/17 - 07/18

Experience

Welcome Desk, Student Project House (SPH) — Zürich

Managed entry control, space operations, and supported 3,000+ students.

Duration: 06/22 - 10/22

Teaching Assistant — EPFL, Lausanne

Taught and supported courses on distributed computing, parallel algorithms, and concurrency models for 200+ students, and designed assignments in Scala.

Duration: 02/21 - 08/21

Master Data Intern — Philip Morris International (PMI), Lausanne

Tested the SAP Master Data platform and created comprehensive documentation for managing millions of data points globally.

Duration: 07/19 - 08/19

Projects

Full-Stack Web Solution for Erasmus Student Network (ESN) Zürich

Co-developed a scalable website for 1,500 users, integrating modern UI/UX design principles and Agile practices.

Duration: 08/24 - Present

Graph-Based Neural Dependency Parsing — ETH Zürich

Implemented a dependency parser using BERT tokenizers, a Biaffine Scorer, and maximum spanning tree algorithms.

Duration: 01/23 - 02/23

Textural Line Segment Classification Research — ETH Zürich

Developed a deep learning method to classify 2D image line segments, achieving a 0.842 F1 score on 5000 images for improved 3D reconstruction.

Duration: 09/22 - 01/23

Binary Sentiment Modeling for Tweet Analysis — ETH Zürich

Designed sentiment classification models analyzing 2.5M tweets with 90% accuracy using pre-trained embeddings and algorithmic refinements.

Duration: 04/22 - 08/22

Tech Stack

Contact

Email: voegelijjm@gmail.com

GitHub | LinkedIn