Artificial Intelligence
Nanjing University of Information Science & Technology (NUIST)
ZAID TANNOURI
Casablanca,CAS
Summary
Self-taught AI / Machine Learning Engineer with two years of formal AI studies at NUIST, China, and continued independent mastery. Ranked Top 5 globally in an active Kaggle NLP classification competition, demonstrating real-world model building and optimization skills. Passionate about creating scalable AI systems and founding a SaaS leveraging machine learning.
Education
Skills
- ML / AI / Deep Learning: NLP (Transformer models, sequence classification, tokenization), Computer Vision (CNNs, object detection), Contrastive Learning, Fine-Tuning, RAG, LoRA / QLoRA
- Frameworks & Tools: PyTorch, TensorFlow, Hugging Face Transformers, Scikit-learn, NumPy, Pandas
- Deployment / MLOps: Docker, FastAPI, Streamlit, Git, Linux
- Other Tools / Practices: Data Augmentation (back-translation, contextual), Embeddings, Boosting Models (LightGBM / XGBoost), Vector Databases
Accomplishments
- Jigsaw – Agile Community Rules (NLP Classification, 2025)
- - Achieved Top 5 (score 0.927 / Prize Contender) among thousands of participants
- - Designed a dual-head model (Category + Misconception) using Qwen3-Embedding-4B + 4-bit QLoRA
- - Combined deep embeddings with boosting models for final predictions
- - Implemented supervised contrastive and ArcFace losses to improve generalization
- - Built robust data augmentation pipeline (back-translation, contextual)
- MAP – Charting Student Math Misunderstandings
- - Ongoing participation in a multi-task NLP competition
- - Prototyping retrieval-augmented learning, embedding optimization, multi-modal techniques
LANGUAGES
Arabic (Native)
English (Fluent)
Chinese (Intermediate)
Timeline
Artificial Intelligence
Nanjing University of Information Science & Technology (NUIST)