Sofiane Hibbah Falaki

• Developed a full anomaly detection pipeline to evaluate instant SEPA payments in near real time, from raw data ingestion to online inference. • Engineered temporal and behavioral features, vectorized device metadata using TF-IDF, and reduced dimensionality with truncated SVD. • Conducted iterative model selection to optimize detection while keeping false positive rates below defined thresholds. • Industrialization: implemented daily orchestration with Airflow, containerized and deployed the FastAPI microservice on Kubernetes, exported models to ONNX format.

• Conducted an in-depth study of theorems in fundamental mathematics, carrying out a rigorous analysis of their proofs. • Developed proof algorithms, translating mathematical reasoning into structured logic. • Produced scientific writing in LaTeX in accordance with academic standards. • Prepared a research dissertation combining proofs, algorithms, and clear, well-structured mathematical illustrations.