Hi, I'm Nafe Abubaker
Computer Engineering Graduate
Computer Engineering graduate from Birzeit University. Trajectory: backend (.NET/MongoDB at ASAL) → data engineering (Python/Prefect ETL pipelines at ASAL) → data science & ML (IBT bootcamp + capstone). Focused on AI/ML and open to offers.
My Education / Experience
Birzeit University
2020-2025Bachelor of Computer Engineering
Relevant coursework: Operating Systems, Database Systems, Data Structures, Object-Oriented Programming, Embedded Systems, Software Engineering.
ASAL Technologies
Sep 2024 - Feb 2025Backend Engineering Training Program
Developed RESTful APIs using C#/.NET Core, integrated MongoDB for database operations, and collaborated with cross-functional teams to build scalable backend solutions.
ASAL Technologies
Feb 2025 - Jul 2025Data Engineering Training Program
Built ETL pipelines with Python/Prefect, processed NDJSON/CSV data from AWS S3, implemented data validation with Pydantic, and created aggregation reports with MongoDB and Pandas dashboards.
IBT Learning
Jan 2026 – April 2026Data Scientist & Machine Learning Trainee
Ongoing intensive remote bootcamp covering Python for data analysis, SQL, and machine learning fundamentals, with hands-on practice in data cleaning, exploratory data analysis (EDA), and data visualization using Python.
My Skills
Technical Skills
Soft Skills
Communication
Clear articulation of ideas and active listening
Problem Solving
Analytical approach to complex challenges
Teamwork
Collaborative and supportive team player
Time Management
Efficient prioritization and deadline adherence
My Projects
PalStoreHub
A full-stack, multi-vendor e-commerce platform that lets Palestinian entrepreneurs spin up their own online stores and list products from a single hub. Also allows tourists to discover various stores in Palestine. Used PATCH instead of PUT across all update endpoints so the frontend can update a single field — like a product price — without re-sending the entire object.
Traffic Light Management & Violation Detection
Capstone project processing drone footage with YOLOv8 + SORT tracking to detect vehicles and traffic-light states across a 4-intersection view. Achieved 91.7% mAP@0.5 on YOLOv8 vehicle detection and improved throughput by ~0.25 cars/s vs fixed-timing baselines (simulation). Also implemented red-light violation detection with stop-line logic and best-frame plate capture + OCR. FastAPI backend, React dashboard, MongoDB. Next time I'd start with 4 fixed cameras (one per intersection) instead of a drone view — easier to deploy in the real world.
Early Warning for High School Dropout Prediction
End-to-end ML project on the HSLS:09 longitudinal dataset (~15.9k students, 5.5:1 class imbalance). Cleaned data, ran EDA, engineered features, and compared Logistic Regression, Random Forest, XGBoost, and LightGBM. Optimized for recall over F1 because missing an at-risk student costs more than a false alarm. Tuned XGBoost at threshold 0.30 catches ~86.5% of actual dropouts; SHAP used to surface top predictors. Next time I'd spend more time on feature selection across the 3,000+ raw columns before modeling.