I am an AI Master's student at the University of Zurich with a background in Statistics and Data Science from LMU Munich. My work sits at the intersection of robust data engineering and advanced machine learning, focusing on Reinforcement Learning, and Time-Series analysis.
- Working at Swissgrid as a Data Engineer, building automated ETL pipelines for energy market transparency.
- Researching RL with a focus on safety-critical domains.
- Exploring Meta Learning and Multi-Modal AI agents.
| Category | Tools & Technologies |
|---|---|
| Languages | Python, Java, SQL |
| AI / ML | PyTorch, Reinforcement Learning (SPMI, DDQN), NLP (RAG), LSTMs/GRUs |
| Data Engineering | ETL Pipelines, API Integration (ENTSO-E, Wikidata) |
| Visualization | Grafana, Weights & Biases (WandB) |
| DevOps/Tools | Git, Docker, RESTful APIs, CLI Development |
- Developed a parallelized variant of the SPMI algorithm for safety-critical Reinforcement Learning.
- Implemented Gaussian Process-based model selection and UCB sampling to improve exploration efficiency.
- Built a flexible framework in Python supporting both tabular and neural network policies.
- A hybrid QA system combining Symbolic Reasoning (SPARQL) and Latent Space Inference (TransE).
- Uses the Wikidata Knowledge Graph to resolve complex natural language queries about the film industry.
- Features an end-to-end pipeline for entity disambiguation and vector-based similarity search.
- Bachelor Thesis: Comparative analysis of LSTMs vs. GRUs for detecting market volatility.
- Integrated a Double Deep Q-Network (DDQN) agent to simulate trading strategies in backtesting environments.
- Utilized WandB for hyperparameter sweep optimization across 1000+ US companies.
- LinkedIn: linkedin.com/in/yourprofile
- Languages: German (Native), English (C1)