J User

Developed a Variational Autoencoder for building typology extraction and automated structural model generation for seismic analysis, achieving significant speed improvements in city-scale risk simulations.

Led the development of a real-time tracking dashboard and a GPS-based driver scoring engine. Directed national-scale data science projects, including speed mapping and road risk analytics, resulting in significant insights for driver safety and risk management.

Developed a GIS platform to scrape and visualize housing prices across 8,000 Spanish towns, beta-tested at the Urban Council of Seville.

Led a project utilizing Python and KUKA robotics for robotic prototyping, resulting in a structure permanently exhibited at ETH Zurich campus.

Trained a neural network to predict the seismic response of reinforced concrete buildings, achieving high accuracy and generating synthetic structural models for analysis.

Designed dashboard visualizations for an EU-wide education readiness tool, utilized across European schools for policy and institutional reporting.

Built a Variational Autoencoder (VAE) to extract building typologies from ~1M cadastral records. Automated structural model generation for nonlinear seismic analysis using OpenSees for ~4k buildings.

Trained a neural network to predict seismic response of RC buildings; achieved accuracy 97.65% predicting stress-deformation curves. Generated 12k+ synthetic structural models and automated SAP2000 nonlinear analyses.

Postgraduate studies focused on Computation & Architecture, ranked top 10 worldwide.