Projects

From data-driven analysis to full-stack applications and cutting-edge AI integrations. Here is a detailed look at some of the projects I'm most proud of.

AI-Powered Language Learning Game

Solo Developer / AI Engineer

Developed a 2D top-down "escape room" game in Unity designed to make language learning interactive and fun. The core feature is the integration of Google's Vertex AI (Gemini API) to dynamically generate questions and prompts for in-game puzzles. Players interact with the AI to solve riddles and unlock chests, creating a unique and personalized learning path with every playthrough. The AI backend was built with Python.

Technologies: Unity, C#, Python, Google Vertex AI (Gemini API)

AI Language Learning Game Screenshot

Hierarchical University FAQ System

Team Member / Software Developer

As part of a four-person team, I co-developed a comprehensive FAQ system for university kiosks using Java and the MVC architecture. My responsibilities included conducting user research through interviews with students and staff to gather requirements and define user personas. I implemented a keyword-based search functionality using Lucene and contributed to the overall system design, which routed unanswered questions to university staff.

Technologies: Java, MVC, Lucene, UML, Git, JUnit

FAQ System UI Screenshot

Full-Stack Android Chat Application

Independent Full-Stack Developer

Independently designed and built a complete real-time chat application for Android from the ground up. Leveraging Firebase for the backend, I implemented key features including secure user authentication, real-time message synchronization with cloud database, online presence indicators, and offline data caching. The UI was built with XML following Material Design principles, focusing on a clean and intuitive user experience. This project demonstrates my ability to handle both front-end and back-end development.

Technologies: Java, XML, Firebase (Authentication, Realtime Database)

Android Chat App Screenshot

Ultra Marathon Runner Performance Analysis

Independent Data Analyst

Conducted an independent data analysis project on a Kaggle dataset of over 200,000 ultra-marathon race records. Using Python and its data science libraries (Pandas, Matplotlib), I preprocessed the data and applied K-means clustering to identify distinct runner segments based on performance and demographics. The findings, which highlighted performance patterns across different race categories and age groups, were compiled into a detailed report.

Technologies: Python, Pandas, Matplotlib, Scikit-learn, K-means Clustering

Data Visualization Chart