My full collection of projects, experiments and software I've built over the years, ranging from machine learning and embedded systems to software engineering and infrastructure.
A full-stack finance application built with React, Django REST Framework and PostgreSQL, developed with a computer vision pipeline to extract transactional data from bank statements. I explored both the application side of personal finance and turning photographed or scanned statements into structured financial data.
A machine learning project comparing different approaches to automating waste classification across nine categories. I worked through the full process from dataset cleaning to training, fine tuning and evaluating classical models, custom CNNs and transfer learning implementations. The project compared classical models, custom CNNs, Azure Custom Vision and transfer learning approaches showcasing the strengths and limitations of each method.
A client server accommodation application designed to help students find and compare places to live. The system combined accommodation data with weather, local crime, postcode and distance information through a Java REST API and desktop client.
An Arduino parcel lockbox designed to secure deliveries when a homeowner is away. I built the firmware and control logic around parcel detection, PIN/RFID authentication, alarm lockout, power loss recovery and a state machine governing the box's behaviour.
A C# desktop application for managing stock in commercial kitchens, with user based access, batch level expiry tracking, automated email alerts, PDF reporting and audit logging. The system modeled different responsibilities across kitchen staff and tracked expiry at the individual delivery batch level.
A multimodal chatbot that can identify landmarks from images, answer questions from a knowledge base and reason about facts provided by the user. I combined image classification, information retrieval, first order logic, AIML and voice input to explore several different approaches to conversational AI within one system.
An experiment in using spiking neural networks to reconstruct clean speech from noisy audio. I built a 302-neuron recurrent reservoir in NEST, inspired by the C. elegans connectome, and trained a readout to convert the resulting spike activity back into an audio signal.
A project constantly evolving using Proxmox environment for Linux, virtualisation, networking, remote access, self-hosting and application infrastructure.
A console-based C++ application using libcurl and RapidJSON to retrieve weather, air quality, historical and location data from Open-Meteo services. GitHub ↗
An individual examination of a raw disk image using Autopsy, The Sleuth Kit, ALEAPP and iLEAPP, including artefact extraction and investigative reporting.
A C++ graphical calculator interface built with SDL2 and SDL2_ttf, including event handling and explicit resource cleanup.
An Arduino accessibility project using six switches to represent a Braille cell and map combinations to characters. My first ever embedded systems project.
Achieved first place by completing a timed interactive cybersecurity challenge.
Worked in a team on realistic technical and business scenarios under tight time constraints during the Bank of America spring insight program.
Completed a technology and financial operations simulation covering banking operations, market infrastructure, risk management under pressure.
A group hackathon project where we developed a budgeting website using Capital One's API.