[nom]

Computer Vision Engineer (Applied R&D)

Email: [e-mail] — Phone: [téléphone] — AGADIR, Morocco

GitHub : [lien] E. — LinkedIn : [lien] [nom] — Kaggle : [lien] E.

Profile

Mathematics graduate with a strong analytical background, currently working in computer vision and deep learning through applied, project-based learning. Experience in designing, training, and analyzing vision models under real-world constraints, including dense scenes and partial occlusions. Interested in applied R&D and building practical, reliable computer vision solutions

Education

• Bachelor‘s Degree in Mathematical Sciences and Applications
• Ibn Zohr University - Faculty of Science, Agadir.
• Graduated: July 2021

Experience

• Applied Computer Vision Intern — PSC Agri (AgriTech Startup) On-site — Sep 2025 - March 2026
• Developed plant instance segmentation solutions for dense agricultural scenes with heavy occlusions.
• Adapted an existing YOLO-based detection pipeline to generate crop-level masks using bounding-box-guided segmentation.
• Designed and trained a Mask R-CNN model from scratch and compared its performance with the detection- based approach.
• Built the full computer vision pipeline from data annotation and preprocessing to model training, evaluation, and deployment through a production-ready inference API using FastAPI.
• Published a technical writeup documenting the methodology: kaggle WriteUp
• Technologies: Python, PyTorch, YOLO, Mask R-CNN, OpenCV, FastAPI

Training and Certifications

• WorldQuant University (Remote)
• Deep Learning Fundamentals Lab (Beta Tester) · Aug 2025 - Now
• Hands-on experimentation with deep learning models, focusing on understanding model behavior, lim- itations, and practical deployment constraints.
• Applied AI Lab: Deep Learning for Computer Vision · Oct 2024 - Apr 2025
• Worked on end-to-end computer vision pipelines, including model design, training, evaluation, and failure analysis on real-world vision tasks.
• Applied Data Science Lab · May 2024 - Sep 2024
• Designed data-driven solutions using Python, statistics, and machine learning, with emphasis on repro- ducible workflows and practical problem solving.
• DataCamp
• Data Scientist in Python. · Apr 2024
• Data Analyst in Python. · Feb 2024
• McKinsey & Company – Forward Program. · Nov 2023 - Apr 2024
• Problem solving, Communication for impact, Digital tools for business.

Languages

• Arabic (Native) · English (Professional) · French (Intermediate)

1

Technical Skills

• Programming Languages: Python, SQL, C, C++
• Machine Learning & Computer Vision: Deep Learning, Data Analysis, Image Classification,Segmentation, Object Detection, Transfer Learning, Vision Transformers (ViT)
• Frameworks & Libraries: PyTorch, YOLO, TIMM, OpenCV, scikit-learn
• Data Visualization: Seaborn, Plotly, Dash, Matplotlib
• Tools & Platforms: Git, Jupyter Notebook, PostgreSQL, VS Code, Colab, Kaggle
• Soft Skills: Problem Solving, Time Management, Independent Research, Self-Learning

Projects

• Date Fruit Classification and Detection System (Kaggle) · Aug 2025
• Developed a multi-task deep learning model to classify Moroccan date varieties and ripeness stages (ripe vs unripe).
• Built a YOLOv8-based detection system to localize fruits and provide fruit-level ripeness assessment.
• Combined classification and detection models into a hybrid pipeline for improved agricultural analysis.
• Tools: PyTorch, TIMM, YOLOv8, Swin Transformer, Multi-task Learning
• Corn Leaf Disease Classification (Kaggle) · Jul 2025
• Designed a classification pipeline using ViT and RexNet for maize leaf disease detection.
• ViT achieved higher accuracy than RexNet on real-world evaluation data.
• Integrated Grad-CAM, precision, recall, and custom Python classes for interpretability and evaluation.
• Tools: PyTorch, TIMM, ViT, RexNet, Grad-CAM, Transfer Learning
• Tomato Leaf Disease Detection (Kaggle) · Jun 2025
• Implemented a YOLOv8-based detection model with real-time inference and confidence thresholding.
• Structured reusable Python classes for deployment-ready pipelines.
• Tools: YOLOv8, PyTorch, OpenCV, Data Augmentation, Transfer Learning
• Wildlife Species Image Classification (Kaggle) · May 2025
• Trained a CNN from scratch on African wildlife camera-trap data.
• Achieved 87% accuracy despite severe class imbalance and limited data availability.
• Tools: PyTorch, CNN, Image Preprocessing, GitHub
• Airplane Crash Data Dashboard (GitHub) · Oct 2024
• Analyzed global airplane crash data (1908–2009) and built an interactive dashboard in Dash.
• Visualized safety trends, causes, and improvements across decades.
• Tools: Python, Pandas, Plotly Dash, Seaborn
• Polish Company Bankruptcy Prediction (GitHub) · Oct 2024
• Built a Random Forest model to predict company bankruptcy risk using financial indicators.
• Visualized results and explained decision boundaries for improved interpretability.
• Tools: Python, Pandas, NumPy, scikit-learn, Matplotlib

2