I specialize in building production-ready AI systems, ranging from LLM fine-tuning and GraphRAG pipelines to Computer Vision projects. Currently completing a Double Master's Degree in Web Intelligence and Data Science (Sorbonne Paris Nord & Sidi Mohammed Ben Abdellah University).
| Domain | Stack |
|---|---|
| Machine Learning | |
| LLMs & GenAI | |
| Data Engineering | |
| MLOps & Tools | |
| Web Development |
| Project | Description | Stack Used |
|---|---|---|
| RAG Course Assistant | Intelligent RAG chatbot that processes PDFs, images, and video to help students understand course materials with source citations. | |
| NeMo Text-to-SQL | Fine-tuned Mistral NeMo 12B using a two-phase incremental learning strategy to transform natural language into complex SQL queries. | |
| Pothole Detection | End-to-end pipeline to identify road defects using YOLO. Handles data preprocessing, training, and optimized deployment. | |
| Vision Benchmark | Comparative study of CNNs, Vision Transformers, and YOLO on the HAM10000 dataset for malignant/benign skin lesion classification. | |
| Multimodal CTR | Hybrid Transformer-DCN architecture fusing CLIP visual embeddings with text to predict Click-Through Rates for the WWW2025 Challenge. |