A Community driven review site of Australian Politicians. Designed to give voters an informed descision before voting. Utilises nodeJS, bootstrap, Pug, ChartJS and built on top of the Hackathon Starter project (nodejs Boilerplate).


A sentiment analysis project based on the sentiment140 dataset. This uses a LSTM based recurrent neural network. Input a sentence of at most 16 words and the model will predict if the sentence has a positive or negative sentiment.

I built an algorithm in C to play the game 2048 which basically involves swiping up, down, left or right to match tiles. The algorithm uses breadth first search to evaluate all moves to a certain depth and determine which move will lead to the best score. As this game contains a random element expectimax search would have been a better algorithm however the asignment specification asked for breadth first search. Above is the algorithm in action and below a link to the source code (left) or alturnatively you can run the code in your web browser. (right)

Note: I did not code the 2048 game itself (my code can be found in ai.c)


I presented a poster detailing how we can exploit quantum mechanical properties of light to securely send a one-time pad without relying on traditional symmetric/asymmetric encryption. Click image for download.