Academic Research
Science | Teaching
Building AI Systems for Scientific Discovery
Scientific data are growing faster than our ability to connect and interpret them. My research develops trustworthy AI systems that combine machine learning, property graph knowledge graphs, large language models, and high-performance computing to transform fragmented biological and environmental data into structured, actionable knowledge. By integrating scientific data, computational methods, and domain expertise, I aim to help researchers generate hypotheses, accelerate discovery, and increase the real-world impact of science.
Details
My work integrates modern computational infrastructures with state-of-the-art data science methodologies, including:
High-performance computing (HPC) clusters for large-scale data processing
GPU-accelerated training of deep learning models
Graph and neural network approaches for complex, interconnected data (supervised, unsupervised, and reinforcement learning)
Knowledge graphs and graph databases for structured data integration and semantic modeling
Large language models to support interpretability and applied scientific workflows
Programming and data analysis using R, Python, shell scripting, awk, Cypher, and SQL
These tools allow me to address data-intensive scientific questions while maintaining methodological rigor and transparency.
Public Helmholtz repositories with research code and tools.
Open-source projects and collaborations beyond Helmholtz.
Publications and citation metrics indexed in Web of Science.
My scientific profile and complete publication record.
