[nom]

# [e-mail] · ƒ [téléphone] · [lien] · [nom] · l [lien]

Profile

AI Engineer and Researcher with expertise in deep learning, computer vision, and predictive modeling. Experienced in building data pipelines, time-series analysis, and AI applications for research and production.

Education

Master in Advanced Machine Learning and Multimedia Intelligence · July 2026

Faculty of Science Dhar El Mahraz (FSDM) · Fes, Morocco

Bachelor in Computer Science · July 2024

Faculty of Science and Technology (FST) · Beni Mellal, Morocco

Work Experience

Bertrandt Group · Feb 2026 – July 2026

AI Engineering Intern (Azure AI Services, Redis, RAG Architectures) · Rabat, Morocco

• Designing, building, and deploying a FAQ chatbot to make internal information easier to access.
• Setting up and improving data workflows to ensure relevant and fast responses.
• Working with different types of documents and making their content usable within the system.
• Evaluating system responses and improving performance through practical optimizations, including caching strategies.

Projects

Dynamic Malware Analysis Using an AI Agent and RL · PyTorch | RL | Dueling DQN | CAPEv2

• Development of an AI agent based on Dueling DQN for dynamic malware analysis, integrated with the CAPEv2 sandbox for efficient sample classification.
• Formulation of the problem as an MDP with a 7-action space (5 investigative + 2 terminal) and a 35-dimensional state representation for progressive feature revelation.
• Training on a subset of the WinMET dataset, achieving a 75.4% reduction in analysis steps.
• Article DOI : 10.5281/zenodo.19513267

Multimodal CTR Prediction on MicroLens-1M · PyTorch | CLIP | Transformers | DCN | FuxiCTR

• Extracted 128-dim multimodal embeddings (CLIP ViT-B/32) for 91k items via attention fusion of image/text.
• Built Transformer + DCNv2 CTR model (FuxiCTR) using these embeddings.
• Achieved AUC 0.87 (validation) and 0.9026 (public leaderboard WWW 2025) with only 1 training epoch.

COVID-19 Forecasting and Analysis · Python | Cassandra | TensorFlow | TFT | CatBoost | Streamlit

• Built a scalable Cassandra pipeline for managing COVID-19 time-series data.
• Trained a Temporal Fusion Transformer for 7-day forecasting and a CatBoost model for patient outcome prediction.
• Designed an interactive Streamlit dashboard for visualization and risk analysis.

Technical Skills

Programming Languages: Python, Java, C, JavaScript, Dart, MATLAB

Machine Learning & Deep Learning: Scikit-learn, TensorFlow, Keras, PyTorch, Edge Impulse

Data & Databases: SQL (MySQL), NoSQL (MongoDB, Cassandra), Neo4j, PostgreSQL

Web & Mobile Development: PHP (Laravel), Node.js, Express.js, Spring Boot, Flutter, Streamlit

IoT & Robotics: Raspberry Pi, ESP32, Arduino

Tools & Platforms: Azure, Git, Power BI, CAPEv2 Sandbox

Languages

• Arabic: Native · • French: Good · • English: Good · • Spanish: Conversational

Achievements

RoboFz Robotics Competition – Fes. · 2024

• 1st Prize in the Remote-Controlled Robots category
• 3rd Prize in the Line-Following Robots category