• Developed and applied Polynomial Neural Networks and Physics-Informed Deep Learning to create machine learning-based approximations for complex simulators, significantly reducing runtimes.• Collaborated on AI-driven solutions for climate crisis policy, utilizing advanced ML techniques to inform critical decisions in public infrastructure and resource management.

• Implemented the integration of GANSim, a 3D generative neural network for generating geofacies cubes conditioned on well log facies and global geospatial data, with Ensemble Smoother Multiple Data Assimilation (ES-MDA) to enhance subsurface modeling accuracy.• Refactored GANSim to a fully convolutional architecture, optimizing the model’s latent space representation for greater control and precision in geofacies generation.• Leveraged Azure cloud for efficient AI model training… Show more • Implemented the integration of GANSim, a 3D generative neural network for generating geofacies cubes conditioned on well log facies and global geospatial data, with Ensemble Smoother Multiple Data Assimilation (ES-MDA) to enhance subsurface modeling accuracy.• Refactored GANSim to a fully convolutional architecture, optimizing the model’s latent space representation for greater control and precision in geofacies generation.• Leveraged Azure cloud for efficient AI model training, inference, and result analysis.• Developed models using Python, TensorFlow, and PyTorch, with a focus on optimization for specific applications.• Managed data preparation, performed result analysis, and authored technical reports. Show less

• Developed and implemented advanced Machine Learning models to effectively predict and analyze employee absenteeism within a 10,000+ workforce, utilizing Python, scikit-learn, SQL, seaborn, Matplotlib, and NumPy.• Successfully deployed models to the cloud and automated their execution, ensuring efficient data processing and analysis. Proficient in leveraging tools such as Digital Ocean, Python, SQL, and Power BI.• Skilled in data modeling, extraction, and maintenance of databases and… Show more • Developed and implemented advanced Machine Learning models to effectively predict and analyze employee absenteeism within a 10,000+ workforce, utilizing Python, scikit-learn, SQL, seaborn, Matplotlib, and NumPy.• Successfully deployed models to the cloud and automated their execution, ensuring efficient data processing and analysis. Proficient in leveraging tools such as Digital Ocean, Python, SQL, and Power BI.• Skilled in data modeling, extraction, and maintenance of databases and data lakes, with expertise in SQL Server, Vertica, and Azure.• Proficient in creating impactful data visualizations using Power BI, Tableau, and Pentaho.• Demonstrated proficiency in generating insightful reports using Report Builder. Show less

• Successfully extracted and processed data to develop score models for the sales department, employing optimal solutions and processes to handle databases with several million rows.• Utilized machine learning models and strategies, including Random Forest, XGBoost, and LightGBM, to target and predict the propensity of sales success.• Conducted in-depth exploratory data analysis and developed business analytics reports using tools such as Excel, Power BI, and QlikView.• Deployed the… Show more • Successfully extracted and processed data to develop score models for the sales department, employing optimal solutions and processes to handle databases with several million rows.• Utilized machine learning models and strategies, including Random Forest, XGBoost, and LightGBM, to target and predict the propensity of sales success.• Conducted in-depth exploratory data analysis and developed business analytics reports using tools such as Excel, Power BI, and QlikView.• Deployed the score models to our internal portal using Docker and Django, ensuring seamless access and utilization by stakeholders. Show less

• As a LHC member I searched how to identify dark matter signatures at the Tracker System of the CMS detector.• From the charged particle hits at the Tracker System, I implemented a Combinatorial Kalman Filter (CKF) algorithm that searches and reconstructs which hits belong to which particle and fits its trajectory through chi^2.• The theoretical formulation of the problem was conceived in cylindrical coordinates, also the fitting, and the dynamical equations of a charged particle… Show more • As a LHC member I searched how to identify dark matter signatures at the Tracker System of the CMS detector.• From the charged particle hits at the Tracker System, I implemented a Combinatorial Kalman Filter (CKF) algorithm that searches and reconstructs which hits belong to which particle and fits its trajectory through chi^2.• The theoretical formulation of the problem was conceived in cylindrical coordinates, also the fitting, and the dynamical equations of a charged particle embedded in a transverse magnetic field were solved using 4th and 5th order Runge-Kutta through SciPy, NumPy and Matplotlib for visualization.Particle collisions' generation, events with dark matter production (FIMP model) with charginos (SUSY theory) as intermediate, through Monte Carlo simulations in the CMSSW framework C++• Exploratory data analysis of data from the simulation to find and characterize the results, by comparison with the state-of-the-art analysis of the area, using Pandas, SciPy, scikit-learn and MatplotlibStudying the frequency distribution of the particles and tracks' parameters, their correlation and resolution of its reconstruction we were able to conclude that dark matter signals could be detected with the increasing energy of the LHC collider, being a possible pioneer discovery of dark matter particles. Show less

• Investigation of charge and energy transport in molecules, describing the problem through quantum mechanics’ formalism and implementing simulations in C++, Python and Fortran.• Using the Hückel extended formalism to define the electronic dynamics of molecules and the Verlet method to describe the nuclear interactions and dynamics, we optimized the non-adiabatic molecular dynamics of Na-(C17-H36)-Na.

• Fundamental mathematical physics research using numerical simulations to solve partial differential equations, relativity andfield theory.• Using new forms of the potential of a field’s theory I found solutions of the dynamical equations implemented with the software Matematica.• The results were presented as posters in two physics’ conferences.

• A study concerning the dependence between galaxies’ metallicity and its age was conducted on a data set of 3,000 galaxies fromthe Sloan Digital Sky Survey (SDSS).• Using the electromagnetic spectrum and the light flux of galaxies, the calculations of indices related to the mean radiation fluxwere implemented (Lick index and D4000), estimating the abundance of ionic metals in the galaxies’s region, and thus its age.SciPy, NumPy, AstroPy were used in the calculations and… Show more • A study concerning the dependence between galaxies’ metallicity and its age was conducted on a data set of 3,000 galaxies fromthe Sloan Digital Sky Survey (SDSS).• Using the electromagnetic spectrum and the light flux of galaxies, the calculations of indices related to the mean radiation fluxwere implemented (Lick index and D4000), estimating the abundance of ionic metals in the galaxies’s region, and thus its age.SciPy, NumPy, AstroPy were used in the calculations and Matplotlib for visualization.• The results were validated with the area’s literature and presented in a physics’ conference as a poster. Show less