I'm a machine learning researcher and bioinformatician specializing in AI x Bio for biosecurity. I'm a Research Scientist at SecureBio, where I build safeguards and evalutions for frontier AI models. In the past, I've built multimodal machine learning / deep learning models for complex biomedical data. I recently completed my PhD at the University of Vienna and the Tomazou lab at the St. Anna Children’s Cancer Research Institute in Vienna, Austria.
-
A benchmark measusing the scientific capabilities of frontier AI models
Together with colleagues at SecureBio, I'm building an evaluation that tests how good AI models are at predicting the outcomes of biological experiments and at generating novel hypotheses. -
DEFT
A multimodal, epigenetics-based deep learning model for enhanced detection of circulating tumor DNA in liquid biopsies - available soon! (Developed by me) -
CellWhisperer
A multimodal AI model that enables intuitive, language-based exploration of single-cell transcriptomics data (Developed together with Moritz Schaefer) -
LIQUORICE
A user-friendly tool for inferring epigenetic signatures from cell-free DNA. (Developed by me)
- Languages & Frameworks: Python, PyTorch, Lightning, W&B, Hugging Face, Snakemake, Pandas, scikit-learn, Inspect AI
- Tools: Docker, Conda, Git, Linux
- See Google Scholar
- Email: <First name>@<Name of the org I'm working at>.org