Projects

Paper Recommendation System
Paper recommendation system based on vector embeddings and user data, featuring a graph-based interface to explore relationships between related scientific papers.

Fast Elo Estimation for Chess
A neural network that plays chess like a human of a chosen rating (600 to about 2400 Elo) and, from the same network, a rating of your own play after every move, with statistical and systematic error bars. Play it in the demo below.

Local LLM with Retrieval-Augmented Generation
Development of a local large language model system using retrieval-augmented generation, combining vector search with tool-based querying and citation-aware responses.

Diphoton Background Modeling (Z→ee)
Ph.D. research project focused on modeling the Z→ee background in diphoton analyses using optimal transport techniques to model fake photons invariant mass distributions.

Diphoton Discrimination with Graph Neural Networks
Exploration of graph neural network methods using reconstruction-level detector information to improve discrimination between diphoton signal and electron background, under constraints from simulation–data mismodeling.

ATLAS Qualification Task – Detector Alignment
Contribution to the ATLAS detector alignment framework within ATHENA, focusing on alignment corrections for the Inner Detector using track–hit residual minimization under strict code review and validation procedures.

Quality Control of Pixel Detector Modules
Development of tools for the visual inspection and quality control of semiconductor pixel detector modules using a linetracing algorithm inspired by the Steger line detection method.

High-Performance Image Stitching
Developed a gradient-based image stitching algorithm optimized for reliability and ultra low memory usage on very high-resolution images (100+ megapixels).

Estimation of the number of multiparton interactions on pp collisions at ALICE using TMVA ROOT
Bachelor’s thesis project using ROOT TMVA to perform feature ranking, decorrelation, and comparison of multiple machine learning models for regression and classification tasks.

Femtoscopy with Monte Carlo Simulations
Monte Carlo study of momentum correlations in high-energy collisions using Pythia, aimed at characterizing the size and shape of the fireball created in heavy-ion collisions

Z-Scan Experiment and Beam Profile Modeling
Experimental optics project involving Z-scan measurements of nonlinear thin films, including data acquisition and mathematical modeling of Gaussian laser beam profiles.