Active Husky is a cross-platform fitness application designed to support group fitness programs by providing users with access to class schedules, fitness resources, and streamlined program management.

The application was built for iOS and Android using React Native and Expo, with a supporting backend infrastructure designed to handle user activity, scheduling, and core application operations.

The system includes a scalable backend built with Node.js, Express.js, and MongoDB, along with an administrative dashboard developed using React, Next.js, and Tailwind CSS for managing application content and fitness program operations.

The project focuses on creating a reliable and efficient platform that improves the organization, accessibility, and management of group fitness services.

This project focused on building an automated pipeline for creating Japanese-to-Mongolian vocabulary flashcards from scanned learning materials.

The system processes cropped textbook images, extracts multilingual text through OCR, and organizes the results into structured data suitable for Anki import.

To address challenges with multilingual OCR, the pipeline separates text by language and writing system, applies Unicode-based filtering to improve extraction quality, and uses spatial information such as Y-axis positioning to correctly associate definitions across different languages.

The final output is a collection of automatically generated Anki flashcards, reducing the manual effort required to create vocabulary study materials.