Filipe Ramalho -Physicsstudent with ML focusBerliner Str. 29,55131 Mainz, Germanyfiliperamalho.de | privat@filiperamalho.de |+49 176 22095047github.com/FilipeRamalho | linkedin.com/in/fmdsr/ProfilMy research work currently involves implementing neural networks (PyTorch) tosolve physical problems. I generally takea data-driven approach (Python, Mathematica) in my work and enjoy workingin interdisciplinary teams. I place great importance on clear documentation and thereproducibility of my results (Docker, CI/CD). In my spare time, I experiment withmodern software development (TypeScript, Kotlin) to expand my skills in cleancode design and user-friendly interfaces.ExperienceWorking Student (Berufsgenossenschaft Holz und Metall)Migrated 2,000+ serversfrom MS SCOM to PRTG.Connection of Grafana toPRTG via an SQL interface.Developed custom PRTGsensors using PowerShell.Supervised trainees.since 11/2023Research Assistant(ZDV Mainz)Performed routine hardware maintenance.Ported websites to modern JavaScript frameworks.Developed Pythonscripts for HPC operations.09/2021 - 10/2023Internship (Institute for NuclearPhysics, Mainz)Designed and soldered circuits forexperiments.Sparked interest in data analysis(evaluating measurement data with Python).06/2018EducationJohannes GutenbergUniversity MainzB.Sc. PhysicsMinor: Computer ScienceThesis: „Neural network implementation of a variational method to determine solutions of theSchrödinger equation“ (Supervisor: Univ.-Prof. Dr. Marc Vanderhaeghen)04/2020 - 03/2026IGS Gerhard Ertl,SprendlingenHigh School Diploma (Abitur)Focus: STEM09/2011 - 03/2020LanguagesGerman (native), Portuguese (native), English (fluent).SkillsListed in descending order of proficiencyML/Data SciencePython, Mathematica, Matplotlib, PyTorch, SQL, Torchquad, NumPyBackend/DevOpsKotlin, PowerShell,Docker, GrafanaWeb DevelopmentTypeScript, AstroTechnical SkillsHardware Diagnostics (WMI), Soldering, Circuit Design