I strongly believe that projects are a win-win activity. They help me learn at an accelerated pace (as against a traditional classroom environment) and (possibly) make a real world impact. Here are some of the projects I’ve actively worked on:
NUS Class Bot - A Telegram bot that automates the attendance taking process in NUS Classrooms. Started out as a fun side project and was later adopted as an official attendance taking tool in CS1101S at NUS. The project was presented at NUS STePS in 2019. Written in Python, the bot uses the Google Sheets API and Redis to store data. As of 2021, it has been used by over 1300 users in NUS over the last two years.
IsoVault - IsoVault is a transactional service builder that adds database read capabilities to static websites by creating easy to embed iframes. We built this prototype as part of Hack for Public Good 2023. Using this service, the Isomer team aims to replace dynamic government websites with static sites, which is estimated to save ~$100 million in hosting and maintenance costs for the government. The project was short-listed and presented to Senior Minister Teo Chee Hean at the demo day for the Hackathon.
Leetcode Zen - After practising on Leetcode for a while, I realized that to best simulate a real world interview - it’s important to not know the problem difficulty, related questions and topics, past submissions history etc. I also wanted to see for myself the limits of building and launching a product with GPT, so I quickly prototyped a browser extension using GPT-4 over a weekend. I also used copy.ai to write the description for the chrome extension store and also used chat-gpt for not only writing the code but also writing descriptions for why certain permissions were required for the extension.
Source-RPC - A module extension that allows the users of Source Academy, an immersive programming environment at the NUS School of Computing, to use Remote Procedure Calls and enable Distributed Computing. This project was completed as part of CS4215 (Programming Language Implementation) at NUS.
Slime RL - A project done as part of NUS CS3244 (Machine Learning). The project used Bayesian Multi-Agent Reinforcement Learning for the Slime Volley environment. The project was completed in a team of six and was presented at NUS STePs, where it won the 1st prize in Machine Learning.
Charitry - A native Android Productivity App that automatically donates to charity whenever the user procastinates. Built using Java, Android Studio and Stripe API at Hack and Roll 2020, a 24-hour hackathon in Singapore.
Webstagram - A static site generator that uses an Instagram Profile as a starting point to create a simple static website. Written in Python with Jinja2, the generator generates HTML, CSS and JS files on every build. This project was built under 24 hours for NTU iNTUition Hackathon 2019.