A news discovery platform that recommends news articles based on user interests and visualises related stories through an interactive article network. Note that the repo is private as the project is still ongoing.

The Midas system is a high-profile initiative responsible for processing financial transactions at scale. This project focuses on Midas Core, the service responsible for receiving, validating and recording the financial transactions. Completed as part of the JPMC Advanced Software Engineering Forage program.

An augmented meal cart for in-hospital use that helps prevent malnutrition in hospital patients by monitoring their nutrient intake. The system captures before and after images/weights of hospital meals, detects and segments foods, estimates portions using depth data, and computes consumed calories, macronutrients and micronutrients using a nutrition database.

A collection of RL policies developed for autonomous wildfire suppression using a fixed-wing drone. The agent learns to navigate a dynamic wildfire environment and optimally deploy fire extinguisher balls. This project explored the impact of different fire spread rates, observation spaces, reward functions and algorithms on agent performance.

A mobile personal training prototype that enables trainers to create workouts, manage trainees and track progress, while allowing trainees to schedule sessions, complete exercises and monitor fitness goals. Features include role-based accounts, multimedia guidance, workout analytics, notifications and cloud database integration.

A full-stack sports centre booking platform with social authentication and cloud deployment that enables members to reserve facilities, manage bookings, find activity partners and report equipment faults. The system also provides staff and administrators with tools for facility management and booking approval.

A distributed storage system that supports concurrent clients and file replication for fault tolerance and scalability.

An analytics dashboard designed to evaluate and visualise the performance of advertising campaigns.

A creative twist on Tetris featuring zero gravity and a multiplayer mode for a fresh strategic experience.

A simulator that recreates the sound of an orchestra performing a given composition, blending musical modelling and procedural audio generation.

A pair of recommender systems trained to predict user ratings for unseen items using collaborative filtering. The first system uses adjusted cosine similarity with top-k neighbour selection and significance weighting. The second system uses matrix factorisation with 100 latent factors.

A fine-tuned ResNet50 model trained to classify brain tumors using axial-view MRI scans. The model achieved an accuracy of 99% on the 4-class tumor classification task. This project explored the impact of data augmentation and transfer learning on model performance.

A collection of deep learning models trained to classify music genres using Mel Spectrograms. The best-performing model (a CNN with batch normalization) achieved an accuracy of 73% on the 10-class genre classification task. This project explored the impact of regularization on model performance.

A K-Means clustering model applied to the text8 dataset to group words based on their semantic relationships using a co-occurrence matrix. This project explored unsupervised learning, word representations and semantic structure discovery in text data.

An XGBoost model trained to predict whether a patient has heart disease based on lifestyle and health indicators. Built as part of a hands-on workshop I delivered on building your first tree-based model, introducing participants to decision trees, ensemble learning and class imbalance.

A recurrent neural network trained to predict Tesla's stock price based on its value over the past 12 days. Built as part of a hands-on workshop I delivered on building your first RNN, introducing participants to sequence modelling, temporal dependencies and time series forecasting.

A convolutional neural network trained to classify images of characters from 'The Simpsons'. Built as part of a hands-on workshop I delivered on building your first CNN, introducing participants to image classification, feature extraction and convolutional layers.

A neural network trained to predict passenger survival on the Titanic using features such as class, sex and family aboard. Built as part of a hands-on workshop I delivered on building your first neural network, demonstrating the fundamentals of data preprocessing, model training and model evaluation.

A decision-driven third-person adventure game where a player embarks on a medieval-themed quest, shaping their fate through choices and consequences.

A third-person zombie shooter where a player battles waves of relentless zombies. Features multiple rooms, each introducing new challenges and enemy spawns.

A fast-paced racing game where a player strives to beat their own lap record. Features both a single-player mode and a mode with AI-controlled opponents.