Data Scientist
Current• Coordinated two data science projects that extracted meaningful insights from complex datasets. The data analysis techniques developed, including the design and implementation of machine learning models, enhanced by a factor of 4 the results obtained in the previous publication.• Engineered the CI/CD pipeline that automated the data analysis, from the data preprocessing to the statistical interpretation of the data.• Developed software for data preprocessing (cleaning, visualization, modeling) and performed statistical analysis on the selected data to test for the presence of a new particle, resulting in a factor 2 improvement.• Optimized the data selection by implementing an algorithm that maximized the chance of detecting the signal produced by a new particle in the LHC data.• Created a tool for statistical interpretation and data visualization, which reduced the time to get the results from more than a week to a single day. Tool used by two research teams.• Built a predictive model able to estimate the misidentification rate of the electron charge in the LHC data using data-driven techniques. Results used by a team of around 40 people.• Automated the analysis of the data taken during the tests of the muon detector, reducing the time to analyze the data from ∼8 hours to ∼1 hour.• Implemented software able to control and monitor the hardware and the data acquisition system of the muon detector, decreasing the manual intervention of the system by 90%.• Documented and maintained tools using a version control system following best practices, with new functionalities added or improved based on the provided feedback and user needs.• Organized weekly group meetings to check the status and evolution of the projects and assign pending tasks. The teams worked in a fast-paced, deadline-driven environment. • Communicated the findings and insights to the stakeholders via weekly oral presentations and written documentation.
