I work as a system architect focused on data analysis and computational modeling, with a particular interest in complex, noisy, and non-stationary systems. My work combines signal processing intuition, non-linear feature engineering, and model robustness, rather than relying solely on off-the-shelf machine learning performance.

I design custom analytical pipelines that transform raw data into structured signal representations, using phase-based and cyclic transformations, normalization strategies, and regularized learning architectures. This approach aims to improve regime separability, detect structural breaks, and reduce overfitting in unstable environments. Beyond individual models, I place strong emphasis on the architecture surrounding experimentation: reproducibility, controlled iteration, ablation, and stability analysis.

To support this work, I build dedicated interfaces and platforms tailored for quantitative researchers, enabling rapid hypothesis testing, comparison of modeling choices, and visualization of regime changes and robustness metrics. These tools are designed to accelerate decision-making rather than optimize isolated scores, and they play a central role in my research workflow.