Crafting intelligent systems with purpose
Software & Intelligent Systems Engineering student at FST Tangier, currently completing a double-degree Master 2 DOSI at Universite de Bretagne Occidentale. I focus on the engineering side of intelligent systems: taking models beyond notebooks and demos into reliable, observable, deployable services. I've worked across healthcare AI, generative AI, geospatial ML, and decentralized platforms — spanning data and model development, APIs, CI/CD, and Docker/Kubernetes. Currently an AI Research Intern at UM6P, applying rigorous evaluation and reproducible pipelines to geospatial ML problems. My interests: AI infrastructure, MLOps/LLMOps, observability, and distributed systems.
“Building production-grade AI systems and CI/CD pipelines with a researcher's rigor.”
Core highlights
- AI Research Intern at UM6P, applying rigorous evaluation and reproducible pipelines to geospatial ML — 78.6%/73.5% accuracy geosite model (1,662 sites) and R²=0.893 Cr(VI) adsorption model
- Built a FastAPI + Whisper + LLM video-CV analysis pipeline at facejob.ma, deployed on AWS ECS with injection-resistant prompt design
- Developed real-time big data pipelines with Spark, Kafka, and Airflow for product recommendations
- Published research on continual learning (FSC-Net) on arXiv
Collaboration
Thrives in cross-functional teams with clear communication and shared ownership of goals.
Problem Solving
Breaks complex ML and DevOps challenges into tractable, deliverable milestones.
Adaptability
Comfortable switching between Unity prototypes, Kubernetes clusters, and AI dashboards.
Communication
Explains technical decisions to stakeholders through workshops, demos, and documentation.