CV

Machine Learning Intern

OceanOS

Developed a marine foundation model using geospatial data to support oceanic intelligence and marine conservation.

• Created geospatial pretraining datasets from NetCDF/Zarr files using Xarray.
• Designed 3D spatiotemporal model architecture for marine data.
• Built Dockerized, multi-GPU training pipeline with cloud support.
• Implemented training failsafes including unit tests and cloud backups.
• Used CodeCarbon for environmental monitoring.
• Achieved 35× training speedup without performance loss.
• Developed end-to-end forecasting model.

Mentor

University of Amsterdam

Mentored a group of 30 second-year students.

Teaching Assistant at FNWI

University of Amsterdam

Led tutorials for ~30 first- and second-year students.

• Programming (various)
• Linear Algebra
• Calculus and Optimisation
• Information Visualization
• Study orientation and career

MSc

University of Amsterdam

Artificial Intelligence

Exchange

University of Cape Town

Exchange program

• Marine Resources and Geography
• Development and Sustainability

BSc

University of Amsterdam

Natural and Social Sciences - Specialization Artificial Intelligence

• Honors program

Reproducing NevIR: Negation in Neural Information Retrieval

• Published at SIGIR 2025.
• Replicated and extended NevIR to assess IR models' handling of negation.
• Found only cross-encoders and listwise LLM re-rankers achieved moderate performance—still below human level.
• Showed poor generalization when fine-tuning across negation datasets.

Geometric Algebra Simplicial Transformer

• Proposed a novel class of equivariant transformers.
• Leveraged simplicial complexes and Clifford algebra to capture full symmetry group equivariance.

Predicting Perceived Attributes of Urban Images

• Investigated the use of GIST features to cluster Amsterdam street scenes.
• Explored how well these clusters capture perceived attributes of urban environments.

Robustness Against Label Noise in Road Defect Severity Estimation

• Researched CNN-based road defect severity estimation and classification.
• Studied impact of label noise and proposed robustness strategies.