Maira E.

I'm a physicist and scientific computing developer with experience in Python, C++, CUDA, AI/LLM evaluation, and data analysis. I combine a strong background in physics and mathematics with practical experience developing computational tools, evaluating AI systems, and solving technical problems. I'm currently at the end of an MSc in Particle Physics at Sapienza University of Rome, where my academic work focuses on experimental physics, astrophysics, and scientific data analysis. **Scientific Computing & Software Development** My projects include: - CUDA N-body galaxy simulation using C++, GPU parallelization, shared memory, and performance benchmarking. - Python-based numerical analysis, simulation, data visualization, and scientific computing. - Game development using Python, Panda3D, Cocos2d-x, and Solar2D, including gameplay mechanics, collision handling, state management, and debugging. - Machine learning projects involving photometric-redshift estimation, model evaluation, and distribution-shift analysis. - Geospatial data processing using Sentinel-2 imagery, spatial validation, and reproducible workflows. **AI, LLM Evaluation & Agentic Systems** I have experience in AI training and evaluation, including: - Evaluating LLM-generated mathematical, scientific, and programming solutions. - Developing and applying evaluation rubrics. - Identifying reasoning errors, hallucinations, methodological weaknesses, and instruction-following failures. - Reviewing prompts, model responses, and technical outputs. - Building and evaluating an agentic clinical-record reconciliation research prototype using structured LLM reasoning, deterministic Python validation, and controlled human-review checkpoints. **Physics Research & Technical Analysis** My research background includes observational astrophysics with JWST data, active galactic nuclei, and experimental work related to dark-photon detection. I have experience in scientific literature analysis, physical modelling, detector instrumentation, data acquisition, uncertainty analysis, and technical documentation. **Scientific Communication & Tutoring** I've worked as a private physics and mathematics tutor since 2020, helping more than 30 students develop conceptual understanding and problem-solving skills. I'm comfortable explaining complex technical ideas, preparing educational materials, and working with scientific documentation and LaTeX. **How I Work** I enjoy understanding a problem from first principles, developing a practical solution, and checking whether the results are physically, mathematically, and technically sound. I'm available for projects in scientific computing, Python/C++ development, AI and LLM evaluation, scientific data analysis, physics-based modelling, technical research, and STEM education.