About Me
I am currently a Postdoctoral Researcher at Idiap Research Institute, in the Neuro-symbolic AI Lab. My work focuses on Information Extraction, Retrieval-Augmented Generation (RAG), Knowledge Graphs and Knowledge Discovery, applied to the Biomedical domain.
After a PhD at INRAE (Toulouse, France) focused on Knowledge Graph applications in metabolomics, I joined the Neuro-Symbolic AI Lab at Idiap (Martigny, Switzerland) in 2022. Combining Knowledge Representation with Natural Language Processing (Information Extraction, RAG), I had the opportunity to apply my research to complex real-world challenges (Antibiotic discovery, Climate risk analysis). Now, I explore how AI methods and Knowledge Graphs can push the boundaries of scientific discovery, especially in the biomedical domain.
I believe in curiosity and collaboration as the core of a healthy work environment. To me, both intense work and informal conversations over a hot coffee (or tea), are the essential ingredients for great projects and breakthrough ideas.
"The whole is greater than the sum of its parts." To me, this has a broad and beautiful meaning. Across scales, from a single cell to entire ecosystems, and to us as a scientific community, interconnected systems naturally generate emergent properties that often exist implicitly, waiting to be revealed. This perspective drives my research: uncovering latent knowledge emerging from the vast landscape of scientific literature.
Experience
Postdoctoral Researcher
Idiap Research Institute, Martigny, Switzerland
- Industry Project Lead MAVERICK (2024 - 2026): Developed a large-scale graph-augmented RAG system for complex, multi-faceted QA across heterogeneous corpora. In collaboration with Basinghall Analytics.
- Industry Project Lead ABROAD (2022 - 2024): Designed an end-to-end hybrid Knowledge Graph–LLM platform for leveraging large-scale scientific literature to support decision-making in antibiotic discovery. In collaboration with Inflamalps.
PhD Student
INRAE ToxAlim, Toulouse, France
Developed FORUM, a large-scale Knowledge Graph integrating scientific literature, life science databases, and metabolic networks for hypothesis generation and predictions of disease biomarkers.
Apprenticeship (MSc in Bioinformatics)
UMR 5088 CNRS, Toulouse, France
Development of circular RNAs and very long intergenic non-coding RNAs detection workflows for analysis of RNA-Seq data.
Publications
Relation extraction in underexplored biomedical domains: A diversity-optimized sampling and synthetic data generation approach
Computational Linguistics
An LLM-based Knowledge Synthesis and Scientific Reasoning Framework for Biomedical Discovery
In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics
FORUM: building a Knowledge Graph from public databases and scientific literature to extract associations between chemicals and diseases
Bioinformatics
Circular ANRIL isoforms switch from repressors to activators of p15/CDKN2B expression during RAF1 oncogene-induced senescence
RNA Biology
Activities
Teaching
Tutorial: Biomedical Knowledge Graphs meet LLM.
Basel Computational Biology Conference, Basel, Switzerland