About Me
I love building whatever ideas come to mind and enjoy turning them into real projects. I'm always working on side projects and try to build out every new idea I have. Coding is something I genuinely enjoy, especially watching my ideas come to life.
Problem-solving and finding creative solutions to complex challenges is where I thrive. With experience across different technologies, I'm eager to keep learning and contribute to meaningful work in an entry-level role.
Outside of coding, I enjoy staying active. I spend time at the gym, play pickleball and volleyball, and like to game in my free time. I'm always open to trying new things and connecting with others who enjoy building and creating.
Skills
Programming Languages
AI & Machine Learning
Cloud Services
Database Management
Software Development & Version Control
Projects

StockSentiment
A full-stack app to analyze and visualize stock sentiment using Reddit posts and financial data. Fetched stock data from yFinance and Reddit posts via API, applied VADER sentiment analysis, and displayed sentiment scores with stock trends on an interactive dashboard. Built with FastAPI (backend) and Next.js (frontend). Deployed backend on Amazon EC2 and frontend on Vercel

AI-Powered Receipt Scanner
During my internship at Squeezee, developed a cross-platform receipt-scanning application using React for the frontend and Firebase for the backend, ensuring seamless functionality across devices. Integrated OCRSpace for text extraction from scanned receipts and parsed and validated OCR results using a Large Language Model (LLM), converting unstructured text into structured, accurate data for storage and future analysis.

Github Vulnerability Scanner
Web app that scans public GitHub repositories for vulnerabilities in the codebase, dependencies, and executable files. It uses GitHub API to fetch repository data and Gemini-2.0 Flash analyze it for potential security issues. The app provides a user-friendly interface for users to input repository URLs and view scan results, including identified vulnerabilities and recommendations to fix these.

BrowserLogger
BrowserLogger is a website that shows everything a browser can reveal about a user without asking for permission, including user actions, device info, location, network details, and more, all through JavaScript and public APIs. It captures things like mouse moves, keystrokes, IP address, screen size, etc.

Hand Tracking and Gesture Detection with AR Smart Glasses
Collaborated with The Digital Camera and Computer Vision Lab at NTU on a real-time hand tracking and gesture recognition system for AR smart glasses, integrating Time-of-Flight (ToF) sensors with Mediapipe for precise 3D hand reconstruction and interactive capabilities. Through optimization efforts, the system's ability to simultaneously process ToF depth data and hand landmark tracking was significantly enhanced, resulting in a 60% FPS improvement from 15 to 24 FPS. Additionally, developed visualization tools for hand tracking and depth data monitoring, which streamlined debugging processes and enabled further system performance optimizations.

Pokernow Analytics Extension
Developed a browser extension that enhances the PokerNow online poker platform by providing real-time analytics and game insights. Implemented features such as tracking player statistics, win rates, and betting patterns to help users make data-driven decisions. Built using JavaScript to ensure seamless integration with the PokerNow interface.

Platformer Game
Created a 2D platformer game using Processing.java, featuring multiple levels, custom physics, character movement, and level design. Developed interactive game mechanics, including obstacle avoidance, jumping, and collectible items. Added background music and sound effects to enhance immersion. Designed challenging gameplay with increasing difficulty to keep players engaged. Focused on smooth performance and responsive controls for an enjoyable player experience.