SibiTrans is an innovative web application designed to bridge the communication gap by serving as an interactive learning platform and real-time translator for Indonesian Sign Language (SIBI). Developed as my final undergraduate thesis project, it integrates a Flask-based web frontend with Google Mediapipe for high-accuracy real-time hand tracking, and a custom-trained Multilayer Perceptron (MLP) deep learning model for sign recognition. The platform goes beyond simple translation; it provides a comprehensive learning ecosystem including a SIBI alphabet dictionary, an interactive learning module, and quizzes to evaluate user proficiency. It also features a built-in tool for customized dataset collection. The future roadmap for SibiTrans focuses on transitioning this robust recognition system into a native mobile application (Android/iOS) to leverage On-Device Machine Learning for even wider accessibility.