Mouheb Hidouri

Building backend ML/AI infrastructure for deZem's anomaly detection and AI reporting platform. Developed FastAPI-based APIs wrapping statistical analysis tools (16 tools), time series forecasting, and anomaly benchmark systems. Currently integrating transformer model outputs into the anomaly detection pipeline for BGMM feature extraction. All services containerized with Docker and deployed via GitLab CI/CD.

• Developed automated ETL and web scraping pipelines using Python, Pandas, and FastAPI, enhancing data collection efficiency by approximately 60%. • Built Dockerized microservices to scale and maintain pipelines, achieving a 35% reduction in downtime. • Implemented machine learning models for HTML content classification, improving accuracy by around 80%.

• Automated IT workflows using GLPI and SCCM, enhancing efficiency by approximately 40%. • Configured and deployed OS/software on over 50 workstations, reducing setup time by around 30%. • Managed IT asset lifecycle, improving asset utilization by approximately 20%.