Ivan Tsz Fung CHAN

is a dedicated Full Stack Developer

with a strong passion for software engineering and a commitment to continuous learning. He holds a First-Class Honours degree in Computer Science from the City University of Hong Kong, where he demonstrated excellence by earning multiple scholarships and receiving regular acknowledgment on the Dean’s List.


Ivan specializes in the development of innovative web and mobile applications, approaching complex challenges with confidence and precision. His focus is on delivering high-quality solutions that have a significant impact on the technology sector. He strongly believes that collaboration is crucial for fostering innovation and driving organizational progress.


Eager to engage in meaningful projects, Ivan is prepared to contribute to the advancement of technology alongside a team of talented professionals.

Skills

React JS

Python

Android Studio (JAVA)

React Native

Angular

NodeJS & ExpressJS & NestJS

MS SQL Server

MongoDB

Git & Github

Postman

Visual Studio

VS Code

Projects Showcase

🌐Web Application

To Do List Web Application

This is a full-stack web application built using MongoDB, ExpressJS, NodeJS, and ReactJS (MERN stack). It supports basic CRUD operations, user authentication using PassportJS & JWT, and image uploading.

Digital Business Card Web Application

This is a full-stack web application built using MongoDb,ExpressJS,NodeJS and ReactJS (MERN stack). It supports registration (with email verification), login, logout, 'change password', 'forget password' (with reset password email), 'Contact site owner'. The Main feature is "Business Card Generation". It provides a form for user to fill in and help generate V-card such that users can download it into their contact book.

📱Mobile Application

Mark Your Loan (on Google Play)

This React Native mobile app simplifies tracking debts and favors between friends. Users can easily record borrowed amounts and payments made on behalf of others, view comprehensive or friend-specific records, and switch between English and Chinese. Data is stored locally using SQLite.

To Do List Android Native Application

This is full-stack android native application built using AndroidStudio, Nodejs, Expressjs, MS SQL server. The API created using nodejs demonstrate how we can perform CRUD in MSSQL database using plain query & stored procedures.The api follows MVC design pattern and preform CRUD on the Task Object in database. Meanwhile, The frontend android application use Retrofit Library for calling the API

To Do List React Native Application

This is full-stack react native application built using ReactNative, Nodejs, Expressjs, MS SQL server. The API created using nodejs demonstrate how we can perform CRUD in MSSQL database using plain query & stored procedures.The api follows MVC design pattern and preform CRUD on the Task Object in database. Meanwhile, The frontend android application use Axios Library for calling the API and performing CRUD operations.

Library Management System

This is a full-stack application built using Android Studio (Java) as frontend and node &express & sql server as backend. This application simulates the daily borrowing and return process of library transactions.

📝 Natural Language Processing (NLP)

CityU Bachelor Degree Assistant [QA bot with Retrieval-Augmented Generation]

The CityU Bachelor Degree Assistant is a Streamlit-based RAG chatbot designed to answer questions using academic PDFs. Powered by LlamaIndex, it leverages local Ollama models (like Qwen3 and Llama3) and Hugging Face embeddings to process data privately. Technologically, it features a small-to-big retrieval pipeline via SentenceWindowNodeParser for precision, alongside a cross-encoder re-ranker. An agentic router directs factual queries to a vector index and broad queries to a summary index. Built for efficiency and uptime, the application implements persistent local storage caching for instant boots and an automated fallback system to maintain stability.

📈Data Science Related

Heart Failure Prediction Using Statistical Machine Learning

This project was conducted to develop and evaluate four statistical machine learning models—Logistic Regression, Random Forest, Gradient Boosting Machine (GBM), and Support Vector Machine (SVM)—for the early detection and prediction of Cardiovascular Diseases (CVDs), specifically heart failure. Given that CVDs are the leading cause of death globally, a Kaggle dataset containing 918 observations and 12 clinical features (such as maximum heart rate, cholesterol, blood sugar, and resting ECG) was leveraged to classify whether a patient has heart disease. Following rigorous preprocessing and hyperparameter tuning, the models successfully identified high-risk individuals, with the Random Forest model achieving the highest diagnostic accuracy at 90.67%.

Study on Hierarchical Clustering

This project implements hierarchical clustering on the wine dataset. It foucuses on studying the differences in hierarchical stuctures generated using single link, complete link and group average. We also make comparison with the clusters formed using k-means and the clusters formed using subset of attributes.

YouTube Spam Classifier

This project adopts different methods in classifying spam comments, which includes Logistic Regression Model,Recurrent Neural Network (LSTM), K-nearest neighbor, Decision Tree Classifier and Support Vector Machine (SVM)

Contact