At FEESC, I develop seismic inversion models and scientific plugins for Petrel, combining applied machine learning research with software development. I work at L3C within UFSC’s Department of Informatics and Statistics, collaborating with LTrace on Petrobras offshore exploration projects.
My recent research developed a physics-informed neural network for joint seismic inversion and automatic wavelet estimation. It combined limited labeled well data with unlabeled seismic data and outperformed a conventional neural-network baseline in low-label experiments; I was first author and presented the work at IEEE IJCNN 2026.
Previously, at LTrace, I developed computer vision and deep learning models for digital rock segmentation. At Heimdall, I redesigned a ZeroMQ messaging system processing millions of tweets per day and built a Bitcoin compliance tool for clustering potentially related wallets and tracing fund flows.
I have a degree in Physics and completed an M.Sc. in Computer Science focused on Machine Learning. I also build AI applications with LangGraph and LangChain, bringing together research and practical software development.
