Zi Fung

🗺️ Soul Maps

A personal media tracker and social sharing platform built with Flask. Track what you read, watch, and listen to and turn it into personalized insights. Soul Maps lets you log books, movies, TV shows, and music, and view your media consumption through dynamic charts and timelines. Add friends, share your media snapshots, and discover new favorites through social sharing.

🏥 Medix: Electronic Medical Record System

A fully functional desktop-based EMR system built in Java. Medix simulates real-world hospital workflows, enabling role-based access for doctors, staff, and patients. It offers complete patient lifecycle tracking, from diagnosis and lab testing to prescriptions and treatment management, all without an external database.

📠 runml (C Mini-Language Transpiler)

A command-line utility written in C11 that compiles and executes programs written in a custom-designed mini-language (.ml). Runml transpiles high-level .ml source code into C11, compiles the result, and executes the output, all within a single utility.

🎧  Spotify Advanced Playback Control Plugin

Currently developing a Spotify plugin designed to enhance playback control through automated actions. Inspired by my daily use of Spotify, I wanted to achieve more flexibility and precision in how playback is managed within the app.

With this plugin, users gain greater control over their listening experience. No matter it’s automatically adding specific songs based on what’s currently playing, or managing the queue with more customisation and ease.

👶🏼 Low Birth Weight Predictor

A statistical modeling project aimed at identifying maternal risk factors associated with low birth weight in newborns. This analysis was completed using R for an applied statistics assignment . The final logistic regression model incorporated interactions and was refined using AIC-based stepwise selection.

💊 Modeling Doctor Complaints with Zero-Inflated Negative Binomial Regression

A statistical modeling project analyzing doctor complaints in emergency services using advanced count regression methods. In this project, I used Zero-Inflated Negative Binomial (ZINB) modeling to handle overdispersed and zero-inflated count data. The model uncovered how workload, income, and gender/residency interactions affect complaint frequency.

🚦 Data Warehouse Design and Implementation for Road Safety Analytics

A real-world data warehousing and analytics project built to support government decision-making on road safety in Australia. The project involved designing and implementing a dimensional model, building a PostgreSQL-based data warehouse, and generating interactive insights using Power BI and association rule mining in Python.

🧩 Graph Database: Road Crash Graph Insights (Neo4j)

A property graph database project focused on fatal road crashes in Australia. Using Neo4j and Cypher query, my report captures complex interactions among crashes, people, vehicles, and locations to reveal actionable insights using graph traversal and pattern matching.