Yeasin Morshed Joy
Software Engineer — Machine Learning
University of Dhaka
Project
Built a machine-learning system to predict home loan approval from demographic, financial, employment, and loan-related data across 78,000 records. Trained and compared XGBoost, CatBoost, LightGBM, and Random Forest classifiers, achieving over 90% accurac
Project
Built a responsive full-stack CRUD platform to manage employee records with search, filtering, and role handling. Engineered an Express.js REST API with comprehensive server-side input validation and JWT-based authentication for secure admin actions. Inte
Project
Developed a full-featured online store with product browsing, product details view, and a real-time responsive UI using React. Built session-persistent shopping cart operations and a secure administrative dashboard for adding, editing, and deleting invent
Project
Built an ML classification pipeline to identify road conditions and surface types across 144,000 road-image/sensor records. Applied ensemble-learning methods (XGBoost, CatBoost, LightGBM) for feature handling, training, and evaluation, reaching 88.89% acc