CarbonSplitGo
PacMan Q-Learning
Playlist Converter
Mastercard Hackathon
Ping-Pong
Pharmacy Website
Quiz Maker
Maze Search
Data Exploration
Gallery App
CarbonSplitGo
PacMan Q-Learning
Playlist Converter
Mastercard Hackathon
Ping-Pong
Pharmacy Website
Quiz Maker
Maze Search
Data Exploration
Gallery App
My Final Year Project for my BCs Computer Science Infrastructure course.
Tools: Swift, XCode, Postgres, AzureVM, MapKit, Figma
GitHub RepositoryThis is an iOS carpooling application made for organisations. The inspiration was my constant struggle when commuting to work, and no viable options available. Some videos which showcase core functionality off CarbonSplitGo. The original research document is over 15000 words, but I am attaching a shortened version which explains my application if interested.
This is a Ping-Pong game running on a stm32f031k6 microprocessor, and being displayed on a 24-bit screen.
Tools: C, Assembly, Python, Keil.
A website in which it is possible to convert YouTube playlists, into Spotify with the help of external APIs.
Tools: Node, JS, W3.CSS, HTML, SQL Server, AzureVM.
GitHub RepositoryDuring my internship at Mastercard I was given an opportunity to participate in a Hackathon. The task was to bridge the gap between online and in-person shopping. To my team I presented the idea of having an app with a map of the shops layout, and the person could see promotional and discounted items. This was a high-fidelity prototype made with Figma, and then presented on a stage in front of a live audience.
Tools: Figma, Canva
My first website ever created. Created a pharmacy shopping website as that allowed me to set up API calls with the database to accomplish basket behaviour.
Tools: PHP, HTML, Bootstrap, MySQL.
GitHub RepositoryProject which uses a less capable Reinforcement Learning technique for a complex game like Pac-Man to discover the limitations. After reducing the possible state action pair space to a manageable ~200,000, the agent was able to learn and achieve 85% maze completion with 15% win rate. This involved various difficulty scaling strategies while learning, and ensuring optimal reward function is set up, while balancing exploration and exploitation using Espilon-Greedy.
Tools: C#, .NET, Avalonia UI
GitHub Repository
As part of my Object-Oriented Programming learning I developed a quiz making app, as such tools allow me to learn information the best. It was a project I did during my early career stages and the UI was not prioritised. The data is simply stored in the local .txt files.
Tools: Java, JFrame.
GitHub RepositoryCreating a Quiz
Solving the Quiz
Autogenerated Quizzes
Maze searching algorithms (DFS, IDS, A*) writen in Prolog. Exploration was done and documented, as this is not a visual project, only the document is attached.
Tools: Prolog
GitHub RepositoryWith the use of datasets found on Kaggle, I did an exploration to understand the reasons for migration. I developed a big idea, and turned it into an analytical project where I did the full date exploration cycle.
Tools: R, QMD
GitHub Repository