Implemented multiple geometrical algorithms in Rust, resultig in up to ~400x speedups over prior implementations
Rewrite a building placement simualtion script from Python to Rust, with ~2500x speedup using parallelism and better algorithms.
Rewrote multiple projects in Rust with web assembly and Python bindings, enabling me to use the a common codebase for both frontend, backend and jupyter notebook demonstrations.
Static Program Analyser implemented in c++ A capstone group project during which I learnt a lot about design patterns and C++
A class project to implement an online code collaboration tool using microservice architecture.
This class taught me about using containers, micro-service architecture and testing through mocking services.
In CS2102, we were tasked to create PSQL queries to satisfy certain requirements. I took this opportunity to attempt test driven development, creating sample datasets using a python script and running the tests using .sh scripts
I took a class in USP called Language, Cognition and Culture. Our deliverable was a report with embedded videos. I created the report in markdown because I was learning to use it, and embedding youtube videos would enable my professor to watch them without needing to download large video files.
In 2020 I took a class on Operations Research. The formulae for our calculations were very tedious and easy to get wrong. I wrote this Python automation tool to aid in verifying that my work was correct.
While taking a class on Engineering Economy, I grew frustrated with the limitations of the hardcopy and excel tables we were given as reference. Therefore, I created a python module to do the calculatiosn and visualisations which were taught in that class.
Luminus was NUS’ learning management system from 2019-2022. Keeping track of the files which I had downloaded and noticing when a Professor had modified or uploaded a new file was difficult, so I wrote a tool to help regularly download new and updated files from the online system.
My first foray into scheduled builds. The dynamic wiki for a game I played was very slow, but by scraping and organising the backing data, I was able to build a static wiki which was much faster. The project uses github actions to update itself daily and scrape any updated data.
Picking self-care tips from a long list requires energy which we might not have, randomly generating them is much less energy intensive
A python package which helps with working with geospatial data from Open data sources and OpenStreetMaps Contains tools to