About
I work on trustworthy machine learning for high-stakes decision making, with healthcare as my main application area. I defended my PhD in Life Sciences (AI for Healthcare) at the Université de Genève in September 2026, in the Data Science for Digital Health group of Prof. Douglas Teodoro. From November 2025 to April 2026 I was a visiting researcher at MIT's Laboratory for Computational Physiology with Prof. Leo A. Celi.
Before my PhD, I obtained an MSc in Electrical Engineering (ir., with Distinction) from Vrije Universiteit Brussel / Université libre de Bruxelles in 2021 and a BEng in Information Engineering from Northwestern Polytechnical University in 2019, and I was a research intern at the Center for Law & Economics, ETH Zurich in 2021.
News
- Oct 2026📄Our survey on retrieval-augmented generation in biomedicine is published in Artificial Intelligence Review.
- Oct 2026📄Two papers in the EMNLP 2026 proceedings: PAINT at the MRL workshop and SeMLI-RAG at the main conference.
- Sep 2026🎓Defended my PhD at the Université de Genève.
- Sep 2026📄Knowledge-enhanced LLMs for multilingual concept normalization published in npj Digital Medicine.
- May 2026🎤Panelist at AI as a Catalyst with Stanford and Threshold Ventures.
- Apr 2026🎭Hosted the first MIT AI Comedy Night: humans against AI on humor.
- Nov 2025✈️Started a research stay at MIT's Laboratory for Computational Physiology.
Research interests
Trustworthy machine learning for high-stakes decision making:
- Transparent human–AI interaction: reliable evidence-based decision support systems.
- Robustness: failure-mode analysis of models deployed in real-world settings.
- Fairness: how model behaviour differs across languages and cultures.
Research highlights
Quality-aware health search
A ranking framework based on discriminative models that automatically detects usefulness, supportiveness and credibility signals across a billion web pages, improving health search results for lay users.
JMIR AI 2024CliniFact
A benchmark for clinical claim verification, where small discriminative models outperformed large generative models at clinical natural language inference.
Scientific Data 2025HealthContradict
A benchmark for biomedical retrieval-augmented generation, showing that in-domain fine-tuning helps models identify the correct context.
npj Digital Medicine 2026PAINT
Led a 17-author study across 16 institutions on disparities in how pain is described in medical text across four languages.
MRL @ EMNLP 2026Publications
First author
- HealthContradict: evaluating biomedical knowledge conflicts in language models
- PAINT: a framework for multilingual pain representation from medical licensing exams to real-world clinical texts
- A dataset for evaluating clinical research claims in large language models
- Online health search via multidimensional information quality assessment based on deep language models
- DS4DH at MEDIQA-Chat 2023: leveraging SVM and GPT-3 prompt engineering for medical dialogue classification and summarization
- DS4DH at TREC Health Misinformation 2021: multi-dimensional ranking models with transfer learning and rank fusion
Co-author
- Retrieval-augmented generation in biomedicine: a survey of technologies, datasets, and clinical applications
- SeMLI-RAG: semantics-guided multimodal late interaction for retrieval-augmented radiology report generation
- Knowledge-enhanced LLMs for multilingual biomedical concept normalization: a multilingual benchmarking and behavioral analysis
- The consensus-based CINEX guideline for reporting clinical information extraction studies
- A scoping review of artificial intelligence applications in clinical trial risk assessment
- An evaluation benchmark for adverse drug event prediction from clinical trial results
- Analysis of eligibility criteria clusters based on large language models for clinical trial design
- Exploring zero-shot cross-lingual biomedical concept normalization via large language models
- A dataset for evaluating contextualized representation of biomedical concepts in language models
- The choice of textual knowledge base in automated claim checking
- Cluster analysis of low-dimensional medical concept representations from electronic health records
Teaching, community & industry
- Teaching assistant for Applied Machine Learning in Healthcare: ChatGPT and beyond, Université de Genève: lectures, slides, exercises and student projects.
- Industry and public-sector collaboration with TriNetX (formerly Clinerion), RisKlick and public health partners on the NLU4EHR, CTxAI and GESICA projects.
- Datathon mentor at MIT Critical Data events with Mayo Clinic (Recognition Award, 2026), Harvard Medical School, Brown University and Bowdoin College.
- Workshop facilitator and panelist at AI as a Catalyst (MIT Critical Data, American Medical Association, Stanford, Threshold Ventures), 2026.
- Founder and host of the MIT AI Comedy Night, a live benchmark of humans against AI on humor, supported by the De Florez Fund for Humor, 2026.
- Second place at the first GATE Challenge, Shanghai, 2025, with Better Buddy (i-Teams, Université de Genève).
- Best Poster, Women in Data Science Conference Zürich, 2025.
- Lightning talks at CUSO Winter School 2025: Human-AI Collaboration, SciMed PhD Conference 2023 and TREC 2021.
- Reviewer for EMNLP 2026, ClinicalNLP @ ACL, npj Digital Medicine, Communications Medicine, Scientific Reports, JAMIA Open and BMC journals.